From ab1f9b810d48b08ba05f03b65d748dff14205cd3 Mon Sep 17 00:00:00 2001
From: nikolasph <nikolas.huelsmann@gmail.com>
Date: Tue, 22 Mar 2016 14:59:43 +0100
Subject: [PATCH] Refactoring
---
Code/FeatExtraction/ClassifMonoView.py | 16 +-
Code/FeatExtraction/DBCrawl.py | 10 +-
Code/FeatExtraction/ExecClassifMonoView.py | 68 +-
Code/FeatExtraction/ExecFeatExtraction.py | 96 +-
Code/FeatExtraction/ExecFeatParaOpt.py | 52 +-
Code/FeatExtraction/ExportResults.py | 40 +-
Code/FeatExtraction/FeatExtraction.py | 59 +-
Code/FeatExtraction/FeatParaOpt.py | 25 +-
.../{ => HelperFunctions}/ExecPlot.py | 21 +-
Code/FeatExtraction/OLD/image_size.py | 57 -
Code/FeatExtraction/OLD/testImage.jpg | Bin 7805 -> 0 bytes
..._03_18-Caltech-ClassLabels-Description.csv | 102 -
.../2016_03_18-Caltech-ClassLabels.csv | 9144 -----------------
...rientaions_8-nbClusters_12-Maxiter_100.csv | 9144 -----------------
...rientaions_8-nbClusters_50-Maxiter_100.csv | 9144 -----------------
.../2016_03_19-FeatParaOpt-HOG-1.csv | 146 -
.../2016_03_19-FeatParaOpt-HOG.csv | 233 -
.../2016_03_20-FeatParaOpt-HSV-H_Bins.csv | 726 --
Code/FeatExtraction/Versions.py | 6 +-
Code/FeatExtraction/__init__.py | 1 -
Code/FeatExtraction/hog_extraction.py | 31 +-
.../hog_extraction_parallelized.py | 33 +-
22 files changed, 290 insertions(+), 28864 deletions(-)
rename Code/FeatExtraction/{ => HelperFunctions}/ExecPlot.py (84%)
delete mode 100644 Code/FeatExtraction/OLD/image_size.py
delete mode 100644 Code/FeatExtraction/OLD/testImage.jpg
delete mode 100644 Code/FeatExtraction/Results-FeatExtr/2016_03_18-Caltech-ClassLabels-Description.csv
delete mode 100644 Code/FeatExtraction/Results-FeatExtr/2016_03_18-Caltech-ClassLabels.csv
delete mode 100644 Code/FeatExtraction/Results-FeatExtr/2016_03_18-Caltech-HOG-CellDimension_5-nbOrientaions_8-nbClusters_12-Maxiter_100.csv
delete mode 100644 Code/FeatExtraction/Results-FeatExtr/2016_03_18-Caltech-HOG-CellDimension_5-nbOrientaions_8-nbClusters_50-Maxiter_100.csv
delete mode 100644 Code/FeatExtraction/Results-FeatParaOpt/2016_03_19-FeatParaOpt-HOG-1.csv
delete mode 100644 Code/FeatExtraction/Results-FeatParaOpt/2016_03_19-FeatParaOpt-HOG.csv
delete mode 100644 Code/FeatExtraction/Results-FeatParaOpt/2016_03_20-FeatParaOpt-HSV-H_Bins.csv
delete mode 100644 Code/FeatExtraction/__init__.py
diff --git a/Code/FeatExtraction/ClassifMonoView.py b/Code/FeatExtraction/ClassifMonoView.py
index a193f1f6..dd14a031 100644
--- a/Code/FeatExtraction/ClassifMonoView.py
+++ b/Code/FeatExtraction/ClassifMonoView.py
@@ -1,22 +1,22 @@
#!/usr/bin/env python
-""" MultiClass Classification with MonoView """
+""" Library: MultiClass Classification with MonoView """
# Import built-in modules
-import pandas as pd
+import pandas as pd # For DataFrames
# Import sci-kit learn party modules
-from sklearn.cross_validation import train_test_split
-from sklearn.pipeline import Pipeline
-from sklearn.grid_search import GridSearchCV
-from sklearn.ensemble import RandomForestClassifier
+from sklearn.cross_validation import train_test_split # For calculating the train/test split
+from sklearn.pipeline import Pipeline # Pipelining in classification
+from sklearn.grid_search import GridSearchCV # GridSearch for parameters of classification
+from sklearn.ensemble import RandomForestClassifier # RandomForest-Classifier
# Import own modules
# Author-Info
__author__ = "Nikolas Huelsmann"
-__status__ = "Development" #Production, Development, Prototype
-__date__ = 2016-01-23
+__status__ = "Prototype" # Production, Development, Prototype
+__date__ = 2016-03-25
##### Generating Test and Train Data
def calcTrainTestOwn(X,y,split):
diff --git a/Code/FeatExtraction/DBCrawl.py b/Code/FeatExtraction/DBCrawl.py
index 1312d77a..4588f8c9 100644
--- a/Code/FeatExtraction/DBCrawl.py
+++ b/Code/FeatExtraction/DBCrawl.py
@@ -1,19 +1,19 @@
#!/usr/bin/env python
-""" Functions to generate DB """
+""" Library: Functions to generate databases of images """
# Import built-in modules
-import os # for iteration throug directories
+import os # for iteration throug directories
# Import 3rd party modules
-import pandas as pd # for Series and DataFrames
+import pandas as pd # for Series and DataFrames
# Import own modules
# Author-Info
__author__ = "Nikolas Huelsmann"
-__status__ = "Development" #Production, Development, Prototype
-__date__ = 2016-01-23
+__status__ = "Prototype" # Production, Development, Prototype
+__date__ = 2016-03-25
# ### Function to iterate through given directory and return images paths and classLabels
diff --git a/Code/FeatExtraction/ExecClassifMonoView.py b/Code/FeatExtraction/ExecClassifMonoView.py
index eab4e6b9..76f6d8db 100644
--- a/Code/FeatExtraction/ExecClassifMonoView.py
+++ b/Code/FeatExtraction/ExecClassifMonoView.py
@@ -1,6 +1,6 @@
#!/usr/bin/env python
-""" Script to perform a MonoView classification """
+""" Execution: Script to perform a MonoView classification """
# Import built-in modules
import argparse # for acommand line arguments
@@ -11,17 +11,17 @@ import os # to geth path of the running script
# Import 3rd party modules
import numpy as np # for reading CSV-files and Series
import pandas as pd # for Series and DataFrames
+import logging # To create Log-Files
# Import own modules
import ClassifMonoView # Functions for classification
import ExportResults # Functions to render results
-
# Author-Info
__author__ = "Nikolas Huelsmann"
-__status__ = "Development" #Production, Development, Prototype
-__date__ = 2016-03-10
+__status__ = "Prototype" # Production, Development, Prototype
+__date__ = 2016-03-25
### Argument Parser
parser = argparse.ArgumentParser(
@@ -32,8 +32,9 @@ groupStandard = parser.add_argument_group('necessary arguments:')
groupStandard.add_argument('--name', metavar='STRING', action='store', help='Name of Database (default: %(default)s)', default='Caltech')
groupStandard.add_argument('--feat', metavar='STRING', action='store', help='Name of Feature for Classification (default: %(default)s)', default='RGB')
groupStandard.add_argument('--pathF', metavar='STRING', action='store', help='Path to the features (default: %(default)s)', default='Results-FeatExtr\\')
-groupStandard.add_argument('--fileCL', metavar='STRING', action='store', help='Name of classLabels CSV-file (default: %(default)s)', default='D:\\classLabels.csv')
-groupStandard.add_argument('--fileFeat', metavar='STRING', action='store', help='Name of feature CSV-file (default: %(default)s)', default='D:\\feature.csv')
+groupStandard.add_argument('--fileCL', metavar='STRING', action='store', help='Name of classLabels CSV-file (default: %(default)s)', default='classLabels.csv')
+groupStandard.add_argument('--fileFeat', metavar='STRING', action='store', help='Name of feature CSV-file (default: %(default)s)', default='feature.csv')
+groupStandard.add_argument('-log', action='store_true', help='Use option to activate Logging to Console')
groupClass = parser.add_argument_group('Classification arguments:')
groupClass.add_argument('--CL_split', metavar='FLOAT', action='store', help='Determine the the train size', type=float, default=0.8)
@@ -45,33 +46,50 @@ args = parser.parse_args()
num_estimators = map(int, args.CL_RF_trees.split())
### Main Programm
-
t_start = time.time()
+# Configure Logger
+dir = os.path.dirname(os.path.abspath(__file__)) + "/Results-ClassMonoView/"
+logfilename= datetime.datetime.now().strftime("%Y_%m_%d") + "-LOG-CMV-" + args.name + "-" + args.feat
+logfile = dir + logfilename
+if os.path.isfile(logfile + ".log"):
+ for i in range(1,20):
+ testFileName = logfilename + "-" + str(i) + ".log"
+ if os.path.isfile(dir + testFileName )!=True:
+ logfile = dir + testFileName
+ break
+else:
+ logfile = logfile + ".log"
+
+logging.basicConfig(format='%(asctime)s %(levelname)s:%(message)s', filename=logfile, level=logging.DEBUG, filemode='w')
+
+if(args.log):
+ logging.getLogger().addHandler(logging.StreamHandler())
+
# Determine the Database to extract features
-print "### Main Programm for Classification MonoView"
-print"### Info: Database:" + str(args.name) + " Feature:" + str(args.feat) + " train_size:" + str(args.CL_split) + ", GridSearch of Trees:" + args.CL_RF_trees + ", CrossValidation k-folds:" + str(args.CL_RF_CV) + ", cores:" + str(args.CL_RF_Cores)
+logging.debug("### Main Programm for Classification MonoView")
+logging.debug("### Info: Database:" + str(args.name) + " Feature:" + str(args.feat) + " train_size:" + str(args.CL_split) + ", GridSearch of Trees:" + args.CL_RF_trees + ", CrossValidation k-folds:" + str(args.CL_RF_CV) + ", cores:" + str(args.CL_RF_Cores))
# Einlesen von Features
-print "### Start:\t Read CSV Files"
+logging.debug("### Start:\t Read CSV Files")
X = np.genfromtxt(args.pathF + args.fileFeat, delimiter=';')
Y = np.genfromtxt(args.pathF + args.fileCL, delimiter=';')
-print "### Info:\t Shape of Feature:" + str(X.shape) + ", Length of classLabels vector:" + str(Y.shape)
-print "### Done:\t Read CSV Files"
+logging.debug("### Info:\t Shape of Feature:" + str(X.shape) + ", Length of classLabels vector:" + str(Y.shape))
+logging.debug("### Done:\t Read CSV Files")
# Calculate Train/Test data
-print "### Start:\t Determine Train/Test split"
+logging.debug("### Start:\t Determine Train/Test split")
X_train, X_test, y_train, y_test = ClassifMonoView.calcTrainTest(X, Y, args.CL_split)
-print "### Info:\t Shape X_train:" + str(X_train.shape) + ", Length of y_train:" + str(len(y_train))
-print "### Info:\t Shape X_test:" + str(X_test.shape) + ", Length of y_test:" + str(len(y_test))
-print "### Done:\t Determine Train/Test split"
+logging.debug("### Info:\t Shape X_train:" + str(X_train.shape) + ", Length of y_train:" + str(len(y_train)))
+logging.debug("### Info:\t Shape X_test:" + str(X_test.shape) + ", Length of y_test:" + str(len(y_test)))
+logging.debug("### Done:\t Determine Train/Test split")
# Begin Classification RandomForest
-print "### Start:\t Classification"
+logging.debug("### Start:\t Classification")
cl_desc, cl_res = ClassifMonoView.calcClassifRandomForestCV(X_train, y_train, num_estimators, args.CL_RF_CV, args.CL_RF_Cores)
t_end = time.time() - t_start
@@ -81,22 +99,22 @@ df_class_res = pd.DataFrame()
df_class_res = df_class_res.append({'a_class_time':t_end, 'b_cl_desc': cl_desc, 'c_cl_res': cl_res,
'd_cl_score': cl_res.best_score_}, ignore_index=True)
-print "### Info:\t Time for Classification: " + str(t_end) + "[s]"
-print "### End:\t Classification"
+logging.debug("### Info:\t Time for Classification: " + str(t_end) + "[s]")
+logging.debug("### End:\t Classification")
# CSV Export
-print "### Start:\t Exporting to CSV"
+logging.debug("### Start:\t Exporting to CSV")
dir = os.path.dirname(os.path.abspath(__file__)) + "/Results-ClassMonoView/"
filename = datetime.datetime.now().strftime("%Y_%m_%d") + "-CMV-" + args.name + "-" + args.feat
ExportResults.exportPandasToCSV(df_class_res, dir, filename)
-print "### Done:\t Exporting to CSV"
+logging.debug("### Done:\t Exporting to CSV")
# Plot Result
-print "### Start:\t Plot Result"
+logging.debug("### Start:\t Plot Result")
np_score = ExportResults.calcScorePerClass(y_test, cl_res.predict(X_test).astype(int))
-filename = dir + filename
-ExportResults.showResults(filename, np_score)
-print "### Done:\t Plot Result"
+### dir and filename the same as CSV Export
+ExportResults.showResults(dir, filename, args.name, args.feat, np_score)
+logging.debug("### Done:\t Plot Result")
diff --git a/Code/FeatExtraction/ExecFeatExtraction.py b/Code/FeatExtraction/ExecFeatExtraction.py
index c59176cc..3412c8ca 100644
--- a/Code/FeatExtraction/ExecFeatExtraction.py
+++ b/Code/FeatExtraction/ExecFeatExtraction.py
@@ -1,6 +1,6 @@
#!/usr/bin/env python
-""" Script to perform feature parameter optimisation """
+""" Execution: Script to perform feature extraction """
# Import built-in modules
import datetime # for TimeStamp in CSVFile
@@ -9,16 +9,17 @@ import argparse # for acommand line arguments
import os # to geth path of the running script
# Import 3rd party modules
+import logging # To create Log-Files
# Import own modules
-import DBCrawl # Functions to read Images from Database
-import ExportResults # Functions to render results
-import FeatExtraction # Functions to extract the features from Database
+import DBCrawl # Functions to read Images from Database
+import ExportResults # Functions to render results
+import FeatExtraction # Functions to extract the features from Database
# Author-Info
__author__ = "Nikolas Huelsmann"
-__status__ = "Development" #Production, Development, Prototype
-__date__ = 2016-03-10
+__status__ = "Prototype" # Production, Development, Prototype
+__date__ = 2016-03-25
### Argument Parser
@@ -29,7 +30,7 @@ formatter_class=argparse.ArgumentDefaultsHelpFormatter)
groupStandard = parser.add_argument_group('necessary arguments:')
groupStandard.add_argument('--name', metavar='STRING', action='store', help='Select a name of DB, e.g. Caltech (default: %(default)s)', default='DB')
groupStandard.add_argument('--path', metavar='STRING', action='store', help='Path to the database (default: %(default)s)', default='D:\\CaltechMini')
-
+groupStandard.add_argument('-log', action='store_true', help='Use option to activate Logging to Console')
groupRGB = parser.add_argument_group('RGB arguments:')
groupRGB.add_argument('-RGB', action='store_true', help='Use option to activate RGB')
@@ -77,7 +78,6 @@ def boolNormToStr(norm):
### Main Programm
-print "### Main Programm for Feature Extraction ###"
features = ""
if(args.RGB):
features = features + "RGB "
@@ -90,12 +90,32 @@ if(args.SURF):
if(args.HOG):
features = features + "HOG"
-print "Infos:\t NameDB=" + nameDB + ", Path=" + path + ", Features=" + features
+
+# Configure Logger
+dir = os.path.dirname(os.path.abspath(__file__)) + "/Results-FeatExtr/"
+logfilename= datetime.datetime.now().strftime("%Y_%m_%d") + "-LOG-FE-" + args.name + "-" + features.replace(" ", "_").rstrip("_")
+logfile = dir + logfilename
+if os.path.isfile(logfile + ".log"):
+ for i in range(1,20):
+ testFileName = logfilename + "-" + str(i) + ".log"
+ if os.path.isfile(dir + testFileName )!=True:
+ logfile = dir + testFileName
+ break
+else:
+ logfile = logfile + ".log"
+
+logging.basicConfig(format='%(asctime)s %(levelname)s:%(message)s', filename=logfile, level=logging.DEBUG, filemode='w')
+
+if(args.log):
+ logging.getLogger().addHandler(logging.StreamHandler())
+
+logging.debug("### Main Programm for Feature Extraction ###")
+logging.debug("Infos:\t NameDB=" + nameDB + ", Path=" + path + ", Features=" + features)
################################ Read Images from Database
# Determine the Database to extract features
-print "Start:\t Exportation of images from DB"
+logging.debug("Start:\t Exportation of images from DB")
t_db_start = time.time()
@@ -105,31 +125,31 @@ sClassLabels = DBCrawl.getClassLabels(path)
# Get all path from all images inclusive classLabel as Integer
dfImages,nameDB = DBCrawl.imgCrawl(path, sClassLabels, nameDB)
t_db = time.time() - t_db_start
-print "Done:\t Exportation of images from DB in:" + str(t_db) +"[s]"
+logging.debug("Done:\t Exportation of images from DB in:" + str(t_db) +"[s]")
################################ Feature Extraction
-print "Start:\t Features Extraction"
+logging.debug("Start:\t Features Extraction")
### Setup RGB
if(args.RGB):
- print "RGB:\t Start"
+ logging.debug("RGB:\t Start")
t_rgb_start = time.time()
# Infos
- print "RGB:\t NumberOfBins=" + str(args.RGB_Bins) + ", MaxColorIntensity=" + str(args.RGB_CI) + ", Norm=" + boolNormToStr(args.RGB_NMinMax)
+ logging.debug("RGB:\t NumberOfBins=" + str(args.RGB_Bins) + ", MaxColorIntensity=" + str(args.RGB_CI) + ", Norm=" + boolNormToStr(args.RGB_NMinMax))
# Extract Feature from DB
rgb_feat_desc,rgb_f_extr_res = FeatExtraction.calcRGBColorHisto(nameDB, dfImages, args.RGB_Bins, args.RGB_CI, args.RGB_NMinMax)
t_rgb = time.time() - t_rgb_start
- print "RGB:\t Done in: " + str(t_rgb) + "[s]"
+ logging.debug("RGB:\t Done in: " + str(t_rgb) + "[s]")
### Setup HSV
if(args.HSV):
- print "HSV:\t Start"
+ logging.debug("HSV:\t Start")
t_hsv_start = time.time()
h_bins = args.HSV_H_Bins
@@ -138,48 +158,48 @@ if(args.HSV):
histSize = [h_bins, s_bins, v_bins]
# Infos
- print "HSV:\t HSVBins=[" + str(h_bins) + "," + str(s_bins) + "," + str(v_bins) + "], Norm=" + boolNormToStr(args.HSV_NMinMax)
+ logging.debug("HSV:\t HSVBins=[" + str(h_bins) + "," + str(s_bins) + "," + str(v_bins) + "], Norm=" + boolNormToStr(args.HSV_NMinMax))
# Extract Feature from DB
hsv_feat_desc,hsv_f_extr_res = FeatExtraction.calcHSVColorHisto(nameDB, dfImages, histSize, args.HSV_NMinMax)
t_hsv = time.time() - t_hsv_start
- print "HSV:\t Done in: " + str(t_hsv) + "[s]"
+ logging.debug("HSV:\t Done in: " + str(t_hsv) + "[s]")
### Setup SIFT
if(args.SIFT):
- print "SIFT:\t Start"
+ logging.debug("SIFT:\t Start")
t_sift_start = time.time()
boolSIFT = True
- print "SIFT:\t Cluster=" + str(args.SIFT_Cluster) + ", Norm=" + boolNormToStr(args.SIFT_NMinMax)
+ logging.debug("SIFT:\t Cluster=" + str(args.SIFT_Cluster) + ", Norm=" + boolNormToStr(args.SIFT_NMinMax))
sift_descriptors,sift_des_list = FeatExtraction.calcSURFSIFTDescriptors(dfImages, boolSIFT)
sift_feat_desc,sift_f_extr_res = FeatExtraction.calcSURFSIFTHisto(nameDB, dfImages, args.SIFT_Cluster, args.SIFT_NMinMax, sift_descriptors, sift_des_list, boolSIFT)
t_sift = time.time() - t_sift_start
- print "SIFT:\t Done in: " + str(t_sift) + "[s]"
+ logging.debug("SIFT:\t Done in: " + str(t_sift) + "[s]")
### Setup SURF
if(args.SURF):
- print "SURF:\t Start"
+ logging.debug("SURF:\t Start")
t_surf_start = time.time()
boolSIFT = False
- print "SURF:\t Cluster=" + str(args.SURF_Cluster) + ", Norm=" + boolNormToStr(args.SURF_NMinMax)
+ logging.debug("SURF:\t Cluster=" + str(args.SURF_Cluster) + ", Norm=" + boolNormToStr(args.SURF_NMinMax))
# Extract Feature from DB
surf_descriptors,surf_des_list = FeatExtraction.calcSURFSIFTDescriptors(dfImages, boolSIFT)
surf_feat_desc,surf_f_extr_res = FeatExtraction.calcSURFSIFTHisto(nameDB, dfImages, args.SURF_Cluster, args.SURF_NMinMax, surf_descriptors, surf_des_list, boolSIFT)
t_surf = time.time() - t_surf_start
- print "SURF:\t Done in: " + str(t_surf) + "[s]"
+ logging.debug("SURF:\t Done in: " + str(t_surf) + "[s]")
### Setup HOG
if(args.HOG):
- print "HOG:\t Start"
+ logging.debug("HOG:\t Start")
t_hog_start = time.time()
CELL_DIMENSION = args.HOG_CellD
@@ -188,54 +208,54 @@ if(args.HOG):
MAXITER = args.HOG_Iter
NB_CORES = args.HOG_cores
- print "HOG:\t CellDim=" + str(CELL_DIMENSION) + ", NbOrientations=" + str(NB_ORIENTATIONS) +", Cluster=" + str(NB_CLUSTERS) + ", MaxIter=" + str(MAXITER) + ", NbCores=" + str(NB_CORES)
+ logging.debug("HOG:\t CellDim=" + str(CELL_DIMENSION) + ", NbOrientations=" + str(NB_ORIENTATIONS) +", Cluster=" + str(NB_CLUSTERS) + ", MaxIter=" + str(MAXITER) + ", NbCores=" + str(NB_CORES))
# Extract Feature from DB
hog_feat_desc,hog_f_extr_res = FeatExtraction.calcHOGParallel(nameDB, dfImages.values, CELL_DIMENSION, NB_ORIENTATIONS, NB_CLUSTERS, MAXITER, NB_CORES)
#hog_feat_desc,hog_f_extr_res = FeatExtraction.calcHOG(nameDB, dfImages.values, CELL_DIMENSION, NB_ORIENTATIONS, NB_CLUSTERS, MAXITER)
t_hog = time.time() - t_hog_start
- print "HOG:\t Done in: " + str(t_hog) + "[s]"
+ logging.debug("HOG:\t Done in: " + str(t_hog) + "[s]")
-print "Done:\t Features Extraction"
+logging.debug("Done:\t Features Extraction")
-################################ SAVE TO FEATURES DATABASES
-print "Start:\t Save Features to CSV Databases"
+################################ SAVE FEATURES TO CSV DATABASE
+logging.debug("Start:\t Save Features to CSV Databases")
dir = os.path.dirname(os.path.abspath(__file__)) + "/Results-FeatExtr/"
### Classlabels and Description
-OutputfileNameClassLabels = datetime.datetime.now().strftime("%Y_%m_%d") + "-" + nameDB + "-ClassLabels"
+OutputfileNameClassLabels = datetime.datetime.now().strftime("%Y_%m_%d") + "-FE-" + nameDB + "-ClassLabels"
ExportResults.exportNumpyToCSV(dfImages.classLabel, dir, OutputfileNameClassLabels, '%i')
-fileNameClassLabels = datetime.datetime.now().strftime("%Y_%m_%d") + "-" + nameDB + "-ClassLabels-Description"
+fileNameClassLabels = datetime.datetime.now().strftime("%Y_%m_%d") + "-FE-" + nameDB + "-ClassLabels-Description"
ExportResults.exportPandasToCSV(sClassLabels, dir, fileNameClassLabels)
format = '%1.30f'
### RGB
if(args.RGB):
- fileName = datetime.datetime.now().strftime("%Y_%m_%d") + "-" + rgb_feat_desc
+ fileName = datetime.datetime.now().strftime("%Y_%m_%d") + "-FE-" + rgb_feat_desc
ExportResults.exportNumpyToCSV(rgb_f_extr_res, dir, fileName, format)
### HSV
if(args.HSV):
- fileName = datetime.datetime.now().strftime("%Y_%m_%d") + "-" + hsv_feat_desc
+ fileName = datetime.datetime.now().strftime("%Y_%m_%d") + "-FE-" + hsv_feat_desc
ExportResults.exportNumpyToCSV(hsv_f_extr_res, dir, fileName, format)
### SIFT
if(args.SIFT):
- fileName = datetime.datetime.now().strftime("%Y_%m_%d") + "-" + sift_feat_desc
+ fileName = datetime.datetime.now().strftime("%Y_%m_%d") + "-FE-" + sift_feat_desc
ExportResults.exportNumpyToCSV(sift_f_extr_res, dir, fileName, format)
### SURF
if(args.SURF):
- fileName = datetime.datetime.now().strftime("%Y_%m_%d") + "-" + surf_feat_desc
+ fileName = datetime.datetime.now().strftime("%Y_%m_%d") + "-FE-" + surf_feat_desc
ExportResults.exportNumpyToCSV(surf_f_extr_res, dir, fileName, format)
### HOG
if(args.HOG):
- fileName = datetime.datetime.now().strftime("%Y_%m_%d") + "-" + hog_feat_desc
+ fileName = datetime.datetime.now().strftime("%Y_%m_%d") + "-FE-" + hog_feat_desc
ExportResults.exportNumpyToCSV(hog_f_extr_res, dir, fileName, format)
-print "Done:\t Save Features to CSV Databases"
\ No newline at end of file
+logging.debug("Done:\t Save Features to CSV Databases")
\ No newline at end of file
diff --git a/Code/FeatExtraction/ExecFeatParaOpt.py b/Code/FeatExtraction/ExecFeatParaOpt.py
index 4c9e8a34..c3b4a527 100644
--- a/Code/FeatExtraction/ExecFeatParaOpt.py
+++ b/Code/FeatExtraction/ExecFeatParaOpt.py
@@ -1,6 +1,6 @@
#!/usr/bin/env python
-""" Script to perform feature parameter optimisation """
+""" Execution: Script to perform feature parameter optimisation """
# Import built-in modules
import datetime # for TimeStamp in CSVFile
@@ -9,6 +9,7 @@ import os # to geth path of the running script
# Import 3rd party modules
import numpy as np # for arrays
+import logging # To create Log-Files
# Import own modules
import DBCrawl # Functions to read Images from Database
@@ -17,8 +18,8 @@ import ExportResults # Functions to render results
# Author-Info
__author__ = "Nikolas Huelsmann"
-__status__ = "Development" #Production, Development, Prototype
-__date__ = 2016-01-23
+__status__ = "Prototype" # Production, Development, Prototype
+__date__ = 2016-03-25
### Argument Parser
@@ -29,6 +30,7 @@ formatter_class=argparse.ArgumentDefaultsHelpFormatter)
groupStandard = parser.add_argument_group('necessary arguments:')
groupStandard.add_argument('--name', metavar='STRING', action='store', help='Select a name of DB, e.g. Caltech (default: %(default)s)', default='DB')
groupStandard.add_argument('--path', metavar='STRING', action='store', help='Path to the database (default: %(default)s)', default='D:\\CaltechMini')
+groupStandard.add_argument('-log', action='store_true', help='Use option to activate Logging to Console')
groupOpt = parser.add_argument_group('Optimisation arguments:')
groupOpt.add_argument('--feature', choices=['RGB', 'HSV', 'SURF', 'SIFT', 'HOG'], help='Set feature from list (RGB, HSV, ..)', default='RGB')
@@ -88,13 +90,31 @@ para_Cl = [args.CL_split, map(int, args.CL_RF_trees.split()), args.CL_RF_CV, arg
### Main Programm
+# Configure Logger
+dir = os.path.dirname(os.path.abspath(__file__)) + "/Results-FeatParaOpt/"
+logfilename= datetime.datetime.now().strftime("%Y_%m_%d") + "-LOG-FPO-" + args.name + "-" + args.feat + "-" + args.param
+logfile = dir + logfilename
+if os.path.isfile(logfile + ".log"):
+ for i in range(1,20):
+ testFileName = logfilename + "-" + str(i) + ".log"
+ if os.path.isfile(dir + testFileName )!=True:
+ logfile = dir + testFileName
+ break
+else:
+ logfile = logfile + ".log"
+
+logging.basicConfig(format='%(asctime)s %(levelname)s:%(message)s', filename=logfile, level=logging.DEBUG, filemode='w')
+
+if(args.log):
+ logging.getLogger().addHandler(logging.StreamHandler())
+
################################ Read Images from Database
# Determine the Database to extract features
-print "### Main Programm for Feature Parameter Optimisation "
-print '### Optimisation - Feature:' + str(args.feature) + " Parameter:" + str(args.param) + " from:" + str(args.valueStart) + " to:" + str(args.valueEnd) + " in #calc:" + str(args.nCalcs)
+logging.debug("### Main Programm for Feature Parameter Optimisation ")
+logging.debug('### Optimisation - Feature:' + str(args.feature) + " Parameter:" + str(args.param) + " from:" + str(args.valueStart) + " to:" + str(args.valueEnd) + " in #calc:" + str(args.nCalcs))
-print "### Start:\t Exportation of images from DB"
+logging.debug("### Start:\t Exportation of images from DB")
# get dictionary to link classLabels Text to Integers
sClassLabels = DBCrawl.getClassLabels(path)
@@ -102,23 +122,24 @@ sClassLabels = DBCrawl.getClassLabels(path)
# Get all path from all images inclusive classLabel as Integer
dfImages,nameDB = DBCrawl.imgCrawl(path, sClassLabels, nameDB)
-print "### Done:\t Exportation of Images from DB "
+logging.debug("### Done:\t Exportation of Images from DB ")
################################ Parameter Optimisation
-print "### Start:\t Feautre Optimisation"
+logging.debug("### Start:\t Feautre Optimisation")
df_feat_res = FeatParaOpt.perfFeatMonoV(nameDB, dfImages, para_opt, para_RGB, para_HSV, para_SIFT, para_SURF, para_HOG, para_Cl)
-print "### Done:\t Feautre Optimisation "
+logging.debug("### Done:\t Feautre Optimisation ")
################################ Render results
-print "### Start:\t Exporting to CSV "
+logging.debug("### Start:\t Exporting to CSV ")
dir = os.path.dirname(os.path.abspath(__file__)) + "/Results-FeatParaOpt/"
filename = datetime.datetime.now().strftime("%Y_%m_%d") + "-FeatParaOpt-" + args.feature
ExportResults.exportPandasToCSV(df_feat_res, dir, filename)
-print "### Done:\t Exporting to CSV "
+logging.debug("### Done:\t Exporting to CSV ")
# Get data from result to show results in plot
+logging.debug("### Start:\t Plot Result")
# Total time for feature extraction and classification
tot_time = df_feat_res.b_feat_extr_time.values + df_feat_res.e_cl_time.values
tot_time = np.asarray(tot_time)
@@ -149,12 +170,13 @@ cl_desc = df_feat_res.c_cl_desc.values
feat_desc = df_feat_res.a_feat_desc.values
store = True
-fileName = dir + datetime.datetime.now().strftime("%Y_%m_%d") + "-" + "Feature_" + args.feature + "-Parameter_" + args.param
+fileName = datetime.datetime.now().strftime("%Y_%m_%d") + "-" + "-FPO-" + args.name + "-" + args.feature + "-" + args.param
# Show Results for Calculation
-ExportResults.showScoreTime(fileName + "-TotalTime.png", store, score, tot_time, rangeX, args.param, feat_desc, cl_desc, 'Results for Parameter Optimisation', 'Precision', 'Total Time (Feature Extraction+Classification)\n [s]')
-ExportResults.showScoreTime(fileName + "-FeatExtTime.png", store, score, feat_time, rangeX, args.param, feat_desc, cl_desc, 'Results for Parameter Optimisation', 'Precision', 'Feature Extraction Time\n [s]')
-ExportResults.showScoreTime(fileName + "-ClassTime.png", store, score, cl_time, rangeX, args.param, feat_desc, cl_desc, 'Results for Parameter Optimisation', 'Precision', 'Classification Time\n [s]')
+ExportResults.showScoreTime(dir, fileName + "-TotalTime.png", store, score, tot_time, rangeX, args.param, feat_desc, cl_desc, 'Results for Parameter Optimisation \n DB: ' + args.name + 'Feat: ' + args.feature, 'Precision', 'Total Time (Feature Extraction+Classification)\n [s]')
+ExportResults.showScoreTime(dir, fileName + "-FeatExtTime.png", store, score, feat_time, rangeX, args.param, feat_desc, cl_desc, 'Results for Parameter Optimisation \n DB: ' + args.name + 'Feat: ' + args.feature, 'Precision', 'Feature Extraction Time\n [s]')
+ExportResults.showScoreTime(dir, fileName + "-ClassTime.png", store, score, cl_time, rangeX, args.param, feat_desc, cl_desc, 'Results for Parameter Optimisation \n DB: ' + args.name + 'Feat: ' + args.feature, 'Precision', 'Classification Time\n [s]')
+logging.debug("### Done:\t Plot Result")
#print 'Les meilleurs parametres sont: ' + str(rf_detector.best_params_)
diff --git a/Code/FeatExtraction/ExportResults.py b/Code/FeatExtraction/ExportResults.py
index e6d1ef2a..1baa3edb 100644
--- a/Code/FeatExtraction/ExportResults.py
+++ b/Code/FeatExtraction/ExportResults.py
@@ -1,6 +1,6 @@
#!/usr/bin/env python
-""" Functions to generate DB """
+""" Library: Functions to export results to CSV or plots """
# Import built-in modules
import os # for iteration throug directories
@@ -17,8 +17,8 @@ from matplotlib.offsetbox import AnchoredOffsetbox, TextArea, HPacker # to gener
# Author-Info
__author__ = "Nikolas Huelsmann"
-__status__ = "Development" #Production, Development, Prototype
-__date__ = 2016-01-23
+__status__ = "Prototype" # Production, Development, Prototype
+__date__ = 2016-03-25
#### Export Features to CSV
def exportPandasToCSV(pandasSorDF, dir, filename):
@@ -52,7 +52,7 @@ def exportNumpyToCSV(numpyArray, dir, filename, format):
#### Rendering of results
### Rendering of Score and Time
-def showScoreTime(filename, store, resScore, resTime, rangeX, parameter, feat_desc, cl_desc, fig_desc, y_desc1, y_desc2):
+def showScoreTime(dir, filename, store, resScore, resTime, rangeX, parameter, feat_desc, cl_desc, fig_desc, y_desc1, y_desc2):
# Determine interpolated functions
f_score_interp = interp1d(rangeX, resScore, kind='quadratic')
f_time_interp = interp1d(rangeX, resTime, kind='quadratic')
@@ -112,7 +112,17 @@ def showScoreTime(filename, store, resScore, resTime, rangeX, parameter, feat_de
plt.title(fig_desc, fontsize=18)
if(store):
- plt.savefig(filename)
+ file = dir + filename
+
+ if os.path.isfile(file + ".png"):
+ for i in range(1,20):
+ testFileName = filename + "-" + str(i) + ".png"
+ if os.path.isfile(dir + testFileName )!=True:
+ plt.savefig(dir + testFileName)
+ break
+
+ else:
+ plt.savefig(file)
else:
plt.show()
@@ -134,15 +144,29 @@ def calcScorePerClass(np_labels, np_output):
### Bar-Plot for score
-def showResults(filename, score):
+def showResults(dir, filename, db, feat, score):
plt.bar(range(0,len(score)), score*100,1)
plt.xlabel('ClassLabels')
plt.ylabel('Score in %')
- plt.title('Resuls of Classification\nCaltech Database')
+ plt.title('Results of ' + feat + '-Classification\n for ' + db + ' Database')
plt.axis([0, len(score), 0, 100])
plt.xticks(range(0,len(score),5))
- plt.savefig(filename)
+ file = dir + filename
+
+ if os.path.isfile(file + ".png"):
+ for i in range(1,20):
+ testFileName = filename + "-" + str(i) + ".png"
+ if os.path.isfile(dir + testFileName )!=True:
+ plt.savefig(dir + testFileName)
+ break
+
+ else:
+ plt.savefig(file)
+
+
+
+
#instead of saving - decomment plt.show()
#plt.show()
\ No newline at end of file
diff --git a/Code/FeatExtraction/FeatExtraction.py b/Code/FeatExtraction/FeatExtraction.py
index f97db5b1..ccc135b7 100644
--- a/Code/FeatExtraction/FeatExtraction.py
+++ b/Code/FeatExtraction/FeatExtraction.py
@@ -1,32 +1,27 @@
#!/usr/bin/env python
-""" Code to Extract all Features from Database """
+""" Library: Code to extract all Features from a Database """
# Import built-in modules
# Import 3rd party modules
-import numpy as np # for numpy arrays
-import cv2 # for OpenCV
-from scipy.cluster.vq import * # for Clustering http://docs.scipy.org/doc/scipy/reference/cluster.vq.html\n",
+import numpy as np # for numpy arrays
+import cv2 # for OpenCV
+from scipy.cluster.vq import * # for Clustering http://docs.scipy.org/doc/scipy/reference/cluster.vq.html\n",
+import logging # To create Log-Files
# Import own modules
-from hog_extraction_parallelized import extractHOGFeatureParallel #For HOG
-from hog_extraction import extractHOGFeature #for HOG computation
+from hog_extraction_parallelized import extractHOGFeatureParallel #For HOG
+from hog_extraction import extractHOGFeature #for HOG computation
# Author-Info
__author__ = "Nikolas Huelsmann"
-__status__ = "Development" #Production, Development, Prototype
-__date__ = 2016-01-23
-
-# ## Functions for Extract Data
-#
-
-# ### Function to calculate the RGBColorHistogram for given Images
+__status__ = "Prototype" # Production, Development, Prototype
+__date__ = 2016-03-25
#### Calculate RGBColorHistograms for all images
-### Points to improve:
# nameDB: Name of the current Image DB
# dfImages: Dataframe with paths to all images - use function imgCrawl
@@ -79,8 +74,8 @@ def calcRGBColorHisto(nameDB, dfImages, numberOfBins, maxColorIntensity, boolNor
sumHist = sum(hist) # Calculates the pixels of each image (images of different resolutions)
if(sumHist==0):
- print "WARNING NORMALIZATION: sumHIST is zero"
- print "image: " + images[1] + "\n"
+ logging.warning("WARNING NORMALIZATION: sumHIST is zero")
+ logging.warning("image: " + images[1] + "\n")
# Normalization
if(boolNormMinMax == False):
@@ -165,8 +160,8 @@ def calcHSVColorHisto(nameDB, dfImages, histSize_, boolNormMinMax):
sumHist = sum(hist) # Calculates the pixels of each image (images of different resolutions)
if(sumHist==0):
- print "WARNING NORMALIZATION: sumHIST is zero"
- print "image: " + images[1] + "\n"
+ logging.warning("WARNING NORMALIZATION: sumHIST is zero")
+ logging.warning("image: " + images[1] + "\n")
# Normalization
if(boolNormMinMax == False):
@@ -206,7 +201,7 @@ def calcSURFSIFTDescriptors(dfImages, boolSIFT):
des_list = []
- print feat + "Keypoints Calculation"
+ logging.debug(feat + "Keypoints Calculation")
#### Feature Detection and Description (Surf):
# Detect (localize) for each image the keypoints (Points of Interest in an image - Surf uses therefore like SIFT corners)
# Pro: SIFT/SURF are scale and rotation invariant!
@@ -217,13 +212,13 @@ def calcSURFSIFTDescriptors(dfImages, boolSIFT):
for images,i in zip(npImages,range(1,len(npImages)+1)):
# Read image
if(float(i)/float(len(npImages))>0.25 and bool_Progress==True):
- print feat + "25% of images processed (Keypoints)"
+ logging.debug(feat + "25% of images processed (Keypoints)")
bool_Progress = False
elif(float(i)/float(len(npImages))>0.5 and bool_Progress==False):
- print feat + "50% of images processed (Keypoints)"
+ logging.debug(feat + "50% of images processed (Keypoints)")
bool_Progress = None
elif(float(i)/float(len(npImages))>0.75 and bool_Progress==None):
- print feat + "75% of images processed (Keypoints)"
+ logging.debug(feat + "75% of images processed (Keypoints)")
bool_Progress = NotImplemented
image = cv2.imread(images[1], cv2.CV_LOAD_IMAGE_COLOR)
@@ -237,7 +232,7 @@ def calcSURFSIFTDescriptors(dfImages, boolSIFT):
kp, des = det.detectAndCompute(image,None)
if des is None:
- print feat + "No Keypoints found in: " + str(images[1])
+ logging.debug(feat + "No Keypoints found in: " + str(images[1]))
if(boolSIFT==True):
desNoKP=np.zeros(shape=(100,128))
else:
@@ -273,12 +268,12 @@ def calcSURFSIFTDescriptors(dfImages, boolSIFT):
else:
descriptors = np.zeros(shape=(size,64), dtype=np.float32)
- #print feat + "Start filling descriptors"
+ #logging.debug(feat + "Start filling descriptors")
for i in range(0,len(des_list)):
descriptors[merker:(merker+len(des_list[i]))] = des_list[i]
merker = merker + len(des_list[i])
- #print feat + "Shape of Descriptors: " + str(descriptors.shape)
+ #logging.debug(feat + "Shape of Descriptors: " + str(descriptors.shape))
return (descriptors,des_list)
@@ -335,7 +330,7 @@ def calcSURFSIFTHisto(nameDB, dfImages, cluster, boolNormMinMax, descriptors,des
# Perform k-means clustering -> creates the words from all describteurs -> this is the (dic) dictionary/vocabulary/codebook
# k: amount of different clusters to build! Will result in a feature length k
- print feat + ":\t Calculation of Dictonary with " + str(int(cluster)) + " Clusters"
+ logging.debug(feat + ":\t Calculation of Dictonary with " + str(int(cluster)) + " Clusters")
dic, variance = kmeans(descriptors, int(cluster), 1)
### 2. Step: encoding/coding/vector quantization(vq) to assign each descripteur the closest "visual word" from dictionary:
@@ -347,20 +342,20 @@ def calcSURFSIFTHisto(nameDB, dfImages, cluster, boolNormMinMax, descriptors,des
# bool for Progressbar
bool_Progress=True
- print feat + ":\t Assign words from Dictonary to each Image"
+ logging.debug(feat + ":\t Assign words from Dictonary to each Image")
npSurfHist = np.zeros((len(npImages), int(cluster)), dtype=np.float32)
for i in xrange(len(npImages)):
# vq: (Encoding) Assign words from the dictionary to each descripteur
words, distance = vq(des_list[i],dic)
if(float(i)/float(len(npImages))>0.25 and bool_Progress==True):
- print feat + ":\t 25% of images processed (Assignments)"
+ logging.debug(feat + ":\t 25% of images processed (Assignments)")
bool_Progress = False
elif(float(i)/float(len(npImages))>0.5 and bool_Progress==False):
- print feat + ":\t 50% of images processed (Assignments)"
+ logging.debug(feat + ":\t 50% of images processed (Assignments)")
bool_Progress = None
elif(float(i)/float(len(npImages))>0.75 and bool_Progress==None):
- print feat + ":\t 75% of images processed (Assignments)"
+ logging.debug(feat + ":\t 75% of images processed (Assignments)")
bool_Progress = NotImplemented
### 3. Step: Pooling - calculate a histogram for each image
@@ -394,8 +389,8 @@ def calcSURFSIFTHisto(nameDB, dfImages, cluster, boolNormMinMax, descriptors,des
sumHist = sum(npSurfHist[i]) # Calculates the pixels of each image (images of different resolutions)
if(sumHist==0):
- print "WARNING NORMALIZATION: sumHIST is zero"
- print "image: " + images[1] + "\n"
+ logging.warning("WARNING NORMALIZATION: sumHIST is zero")
+ logging.warning( "image: " + images[1] + "\n")
# Normalization
if(boolNormMinMax == False):
diff --git a/Code/FeatExtraction/FeatParaOpt.py b/Code/FeatExtraction/FeatParaOpt.py
index b98e139a..2ad00dd7 100644
--- a/Code/FeatExtraction/FeatParaOpt.py
+++ b/Code/FeatExtraction/FeatParaOpt.py
@@ -1,22 +1,23 @@
#!/usr/bin/env python
-""" Function to optimise feature parameters """
+""" Library: Function to optimise feature parameters """
# Import built-in modules
-import time # for time calculations
+import time # for time calculations
# Import 3rd party modules
-import numpy as np # for numpy arrays
-import pandas as pd # for Series and DataFrames
+import numpy as np # for numpy arrays
+import pandas as pd # for Series and DataFrames
+import logging # To create Log-Files
# Import own modules
-import FeatExtraction # Functions for Feature Extractions#
-import ClassifMonoView # Functions for classification
+import FeatExtraction # Functions for Feature Extractions#
+import ClassifMonoView # Functions for classification
# Author-Info
__author__ = "Nikolas Huelsmann"
-__status__ = "Development" #Production, Development, Prototype
-__date__ = 2016-03-14
+__status__ = "Prototype" # Production, Development, Prototype
+__date__ = 2016-03-25
# dfImages: Database with all images
@@ -58,7 +59,7 @@ def perfFeatMonoV(nameDB, dfImages, para_opt, para_RGB, para_HSV, para_SIFT, par
# for-loop from valueStart, step stepwidth, until valueEnde
for valuePara, i in zip(valueArray,arr_Calc):
valuePara = int(valuePara)
- print "### Start:\t FeatureExtraction Nr:" + str(i) + " from:" + str(max(arr_Calc)) + " with " + parameter + "=" + str(valuePara) + " ###"
+ logging.debug("### Start:\t FeatureExtraction Nr:" + str(i) + " from:" + str(max(arr_Calc)) + " with " + parameter + "=" + str(valuePara) + " ###")
# TIME for Extraction
t_extract_start = time.time()
@@ -160,7 +161,7 @@ def perfFeatMonoV(nameDB, dfImages, para_opt, para_RGB, para_HSV, para_SIFT, par
# TIME for Extraction END
t_extract = time.time() - t_extract_start
- print "### Done:\t FeatureExtraction Nr:" + str(i) + " from:" + str(max(arr_Calc)) + " ###"
+ logging.debug("### Done:\t FeatureExtraction Nr:" + str(i) + " from:" + str(max(arr_Calc)) + " ###")
# TIME for CLASSIFICATION
t_classif_start = time.time()
@@ -178,11 +179,11 @@ def perfFeatMonoV(nameDB, dfImages, para_opt, para_RGB, para_HSV, para_SIFT, par
# Begin Classification RandomForest
# call function: return fitted model
- print "### Start:\t Classification Nr:" + str(i) + " from:" + str(max(arr_Calc))
+ logging.debug("### Start:\t Classification Nr:" + str(i) + " from:" + str(max(arr_Calc)))
cl_desc, cl_res = ClassifMonoView.calcClassifRandomForestCV(X_train, y_train, num_estimators, cv_folds, clas_cores)
- print "### Done:\t Classification Nr:" + str(i) + " from:" + str(max(arr_Calc))
+ logging.debug("### Done:\t Classification Nr:" + str(i) + " from:" + str(max(arr_Calc)))
# TIME for CLASSIFICATION END
t_classif = time.time() - t_classif_start
diff --git a/Code/FeatExtraction/ExecPlot.py b/Code/FeatExtraction/HelperFunctions/ExecPlot.py
similarity index 84%
rename from Code/FeatExtraction/ExecPlot.py
rename to Code/FeatExtraction/HelperFunctions/ExecPlot.py
index 7fae2453..27ac5f2e 100644
--- a/Code/FeatExtraction/ExecPlot.py
+++ b/Code/FeatExtraction/HelperFunctions/ExecPlot.py
@@ -1,10 +1,25 @@
-import pandas as pd
-import numpy as np
-import datetime
+#!/usr/bin/env python
+
+""" Script whichs helps to replot results from Feature Parameter Optimisation """
+
+# Import built-in modules
+import datetime # for TimeStamp in CSVFile
import argparse # for acommand line arguments
import os # to geth path of the running script
+
+# Import 3rd party modules
+import pandas as pd # for Series
+import numpy as np # for DataFrames
+
+# Import own modules
import ExportResults # Functions to render results
+# Author-Info
+__author__ = "Nikolas Huelsmann"
+__status__ = "Prototype" # Production, Development, Prototype
+__date__ = 2016-03-25
+
+
parser = argparse.ArgumentParser(
description='This methods permits to execute a multiclass classification with one single view. At this point the used classifier is a RandomForest. The GridSearch permits to vary the number of trees and CrossValidation with k-folds.',
formatter_class=argparse.ArgumentDefaultsHelpFormatter)
diff --git a/Code/FeatExtraction/OLD/image_size.py b/Code/FeatExtraction/OLD/image_size.py
deleted file mode 100644
index 01fad68b..00000000
--- a/Code/FeatExtraction/OLD/image_size.py
+++ /dev/null
@@ -1,57 +0,0 @@
-
-import os as os # for iteration throug directories
-import pandas as pd # for Series and DataFrames
-import cv2 # for OpenCV
-import datetime # for TimeStamp in CSVFile
-from scipy.cluster.vq import * # for Clustering http://docs.scipy.org/doc/scipy/reference/cluster.vq.html
-import numpy as np # for arrays
-import time # for time calculations
-from feature_extraction_try import imgCrawl, getClassLabels
-
-def findmaxDim(npImages):
-
- max_height = 0
- max_width = 0
- heights =[]
- widths = []
- totals = []
- count=0
- poulet=0
-
- for npImage in npImages:
- height, width, channels = cv2.imread(npImage[1]).shape
- heights.append(height)
- widths.append(width)
- totals.append(width*height)
- if width * height > 500000:
- count+=1
- if not(abs(height-200) <200):
- poulet+=1
- print (npImage[1])
- if width > max_width:
- max_width=width
-
- print float(poulet)*100/len(heights)
- # print float(count)*100/len(heights)
- return heights, widths, totals
-
-
-path ='../../03-jeux-de-donnees/101_ObjectCategories'
-
-# get dictionary to link classLabels Text to Integers
-sClassLabels = getClassLabels(path)
-
-# Get all path from all images inclusive classLabel as Integer
-dfImages = imgCrawl(path, sClassLabels)
-
-npImages = dfImages.values
-heights = []
-heights, widths, totals= findmaxDim(npImages)
-heights_ = sorted(list(set(heights)), reverse=True)
-widths_ = sorted(list(set(widths)), reverse=True)
-totals_ = sorted(totals, reverse=True)
-# print (totals_[len(totals_)/2])
-# print ("height", sum(heights)/len(heights), "width", sum(widths)/len(widths) )
-# print heights_
-# print("poulmmet")
-# print widths_
\ No newline at end of file
diff --git a/Code/FeatExtraction/OLD/testImage.jpg b/Code/FeatExtraction/OLD/testImage.jpg
deleted file mode 100644
index d87740af77771b77e2203ccc5bdc04969b20e374..0000000000000000000000000000000000000000
GIT binary patch
literal 0
HcmV?d00001
literal 7805
zcmex=<NpH&0WUXCHwH#VMg|WC4+e(+4;jQ<D>Bm<7<_#hv=|r|I2gd-|7`|mkSaz>
z0LTzZ^n>;OKfoZ!!ML8Wgqcx^fk}{&S&;Gn5e6v+Mn(o^Mn<q>Ap|!ECl?$0|04_z
z0t`${OpMG-%*^autgIkeHWpTP4rWeKF(xiy5phO|4oNl3$Qe7X3kpf8s(Xe{KXF4v
zBdoN-)FPs6#s4D=A}9uf3}66>3L;wv5xvF0!_3IQB*-ktV9)UW0a$R@5X>>s)mL9H
z>+6U(8+R-A_~9QPZ!K+VP5Uk$92uSJb82(axz(rMI;w8we#ABX&YETWGuFQBE|lCj
z(Q0k@+Xp)%AAP;^OJ%EWK~&rEgU7^9|GxUAT>g93+L&v;4L8NOavE*R_w2v5GJBm)
zhReG57xyh%YFwWqY35!%o3kqDQTI8K<J)CcPv4v)x%0{mO&h=S9@g&fO<!3~{qw9n
z%Fu8BtE{R28G3#M%cpKR_wy$E)*oN4*`E*pSM^5a>eH~&>lf_TZe85Iak=5Q8S;YL
z<_Z62Sflxklj%Q$m-V{+{~0Xz|2eyS{@X7H_)i_ZYW{Uv{(pwI6F0^mNznbo(Oeg%
zw{riQ&Gr8oa(cV|DPvLcY31)(R*zLztyv!YzL@1ytLd-QSvz}Yct^%ZEe#J0cU#@8
zI`idgZo}PH2F1s^kDe~N@#t{XmE*~;CKnto&dfJTc~|qF;fj65^Po4{$y~cLLks@B
zQu$;MA9L@_&e)X)ET_L+H*59LaEE2fzIJFRKAI)^OKe7A)RU7b*N-n-J&#)~^?tO8
z#O1K%>aA`EPl&CnK3i4zw4QC@&6?m1QnPN?Ke?B6{ASkYeVmSO>zqxCKVM~@x7%!M
zWoT+C5C5yV84us&>C`T(KN7Tj)_;cAFRBB6P02|6749^1#qZGe&$fGx{(g2Yrq<{5
zi$!O)tXl0G;~CC#c82Nk)?y>o^&gD9)-+2!-6=L{*XgFOX_q$noN6n5c`b1^;k3^-
zp=w>}=Zp8IT=_IDc81;VH&@oZkM~{uqjsuS^{uPNrBu>gXWM+W@?Ecfp{MTn{fmoT
z_U>7>zk1j5!iPD>ZfsW#T4z@Evb@GgI%`|J<~z6JX}ycjmfgLt_FSg1?%&xD>Ko55
zU-a(NzI$h;EPguuERX1n{|xf$KZYtNFFnmBT`%Gi8yneeK26jm{|WE@uX^8=zRJ8Z
zf7&5e{XPC#7~k)}@AU`5!n*$j&CPy3W$$Fu)ckIFh17lD`|sIkE3dA3RmWNT@mHGt
z!tC!c_TrZp2*7i>T%+&Cm7FV6Uy9Cb)V-<OG;^cnQNL$L3s?HCEIMX-=(EvjUYi+Z
zUvxymK0Le?9CP(dcXD{Ln8^#H3Y8CqMZao(Ft5Ki%kF*c$!qz-M|Ypw`~Bjf(7Rpj
zuk_D4hTbY!dZLbP_ve4HA1bbF-ruJgus!bEoW=3pQ$E$XX1&Y1tsfgH|Dbm6zmhMn
zN|)UYC`_FfbR}p0(|xLsHhh|9m9l)!(kUOArL&$hi&j1U)jNOjzxS@3SDg-S7c<?s
z((0O`{k9)d-`i+3OWLofUby*}_{_xgzwW2sNF-}J$L|Y%nz*cGe|2ohs?!IL9xd2(
zW6op0W0PL&Kd+{ndM7LG>FHgsuB}p?_UpujTVdksw3%KjJa1ogeMfqP)v{HO|G0e3
z+uvgTeQs^{#N+WhcStC$dSX}N_3AX6_#LL}g~yE+{jB@7DCO9SACoSwDob@=7jq&$
z<<)f;nd;LsX8w|chc0%8gUtSe3<aoiwoJ2=gM@=lU(z;airy5>YMfNuHTP`S<Dl6=
zPkXXkpQTQdsx)}}B0^*B!^2ya?OZ)m?AW?vI++(zK6pNO^z_S(?1lA_Tl_ZMdR%u_
zHTdPavkCpYu{Y-wMshu#C}exMCyYO+&h_XTF}t1r8BF^>g}0k|t`FR~bKdtpJq`Vk
zn>%m3Oj@4ure7x5{8wrIwA7b7w@b{=`f&C#JIj=9Tq|~Jip^|qEZS|BxH7y{c$1k(
z&P==3O^Yw>`yCkmqEqVHQQ!W#imSBFc}A>%Q05)6WgT}<^Yb6`HI>q3B72SFboYMh
zNu6Q%D=+A{&)rRrCZCvS#Pl>fsrP+a()tw%E8e~S_Vt_Wrf;h@Jz_sovvzKc=AM|3
zJN7l81`ryr#q7p|{k#5MezR>$rT?7TyWg0N684+EnP>b=-qT21ihatD=j90>|1&6j
z{P*^z_1F7ZKle{~_bp!k^PKCImJeUazMb*%<(26fQrjNu?RfLg<kb1qFRz^66A`Su
z;OeUFHx}z3@mrX)Vfo1&)3rQTCVWx3thM4)&+4>2tJQY;tvdfU^2?SRCI%Z?I{o^$
zMBlo2HNHSkSg*)V_4mv~HU93z;z`?_-#(77UcY>Kb<={gk^bupm(}L}YP@aJxKXKF
zUFVUCdsOK^OS?5o_D7sso?q&^m+wXYW%uQ>z1n@-v$Os)EH(;RzdpF=`JcI|xl@hx
zZF`N@{%Sp(^jo@THCK<%!3kMXH+O8Z$r0Mybv1g2UTw(duv+d??JHA$$1_jRjro>Z
z^7yD6Z|$a^(FtZJw>6*KbG3MH=z|`izjtlggq}YOVrT8k4gDB(T1xUi!&Phjv-f4f
zt+)JVXj@un%6xJE8~;m_mxb~D%2XG7!M8wrL*jo1-si{t_m`hZET3{c&+tD(?vud0
zU)zps&7Yk6iPhy+*YEi&wy&L*Uwq^7Pt(~V*ZAV9^<w^mBCO^=gJx1q<eWhL!o2Kb
zGyXFeB&+q^d~)NprrNp`7B{_<g>grH?}tQd=U#vN+nLoh^lf^<`iY;!MW=o%tP{Br
zxbw%B<}GzM*7fdod~Wuip^Sg`K8K2r^WSoZOTRBvZ~t+=-}TS2!}~>>Bg6kRO|9Q#
zT%TC<dwq3c=6{Cv6Jl|Rhxb2VnOU{)cg*+m%bQkj%RfF%`*-{54(-pgWs2%7R)1Zd
za;9jtd(6UDiD@Tqs6Bf3?Vi*h>*{Boa<k_I-Jhww!tj+co6o69g|+hXIuUa&O#flw
z{N4KtkIOcf)3#xyy!A(mnAU%?ekUgq7PVvJ@>3??%mp58e$x~iv^TWB>j`Vz<5N#J
z{F$5_u4KL_-1NEh>(-ctRo@e5IaMTpLWe=)=bKZn^uKDeJk6V>|IJ=-d(7`^CD;Em
z{P2JF=O1TX;lp2R*8cVCXWFEHIp)@5-pzY=?TfS6@%O=Qp{FjUI)SG@?cFEX`htJ1
z-MhajyQ=MfY@bodb|Cn{zRUOi)IYWRA`*V<;!}J3TYCSJ5B^*GbA8HBJ9)K_{~2W7
zz5aWBf7s0zU`G%^aD4jRZ&kPI%ZFFJQzsU#XZq_Weg5><*})yc(|^2P@F96>%!)IQ
z{Z9Ao)xN^nsdb9`m1Ly4a_D1G(TZ}@D-zlLGLG|(CY^k-_fh8}q3LyBYcqZ|F5Y!=
zOKGSj)6?eYZA@30o*z#r7jdnL+~^fQ=iE;5rz_IVF8NtAyZyr=qi&gHQmW^)ds@$J
zPUL#qxY2)`FSz<*ES;-#HaA*x>95X>{h_bF>fQVuVIY=WxH`BynKf+dYOSQH=F6nM
z&M+6-e(l5chqv#Y>Gu`8^y%3#+j5ya$!pt;YGpQCF7%1WzO8rT%F7GSUGiiiXGrTl
zT>tE<MOn{vGxKYPp#j^nO%^R$b8K#=TF5-sFI%P_ee%>QPcPoL=+RnHqpvm@rJ@Ci
zAK1R*FRKtCoWXuq_7|7=><?#uKYaaCA!vWxi}x1-_uOyGzt11<$pkh=`|I1nzpPTm
zKW6^@@Qq`FX8ii>{ude-=N~!$zI}s3{ttv{cv<W^aUb?vyV~{3-orI+ef-S8O|$J=
zHY~o<J!6}ZPF85fLYp@!)~t3JW?L^V*idAhAo6Ob(K^NZ;Z;|2W-j;`T5$W!>OIS*
zDqefPhk1q2x=Zh8>8(tjSNi+NT<-Y%;MBrZPu6XEzWw~p>T9zLr(a9jaM5zZl?eUE
zUM)FUt%<(2g_al1jV?Xgn$~fxaCL@dQjrVTv0KhN-O*UxP-S28&ed;2!lo5_ukn=6
zdghTfHP+~O@9I>=qBqG`Yx1&ot;)F9`fGa5io;>c#iz-e_J~H5T*%YS&wYGu&&Gbe
z<q7-uvR!`dx24j*`R^k$X5I&&qTl!7e}=G6T93eGfBYwY`A>YHqF=rAr~ia^U-uh+
zo?~7Kb_F4V@dMAxZB`ps|I__G{bp|c^?Hl<b)sJ{8zrytd?sgn%VzS?J%4-aPuSF-
z(5Swj`E%dZd#~*3pZ=a=-y>gd!hHSw+A^J&f$MZQy>kBEul(F+pW3+2gL_Z+<adVK
z%e%97GN*phQS%CG{dK0k|Axhqvw;iKVz&Rf@oo3HM&IC5>jS687H(Tm)&2WyfuHF+
z4Ox>y|9ghpFMgW;nQ^XWOhADT|MLpl>(91n{&;zFyK2qVl<!AY@7I~VexJ-^_kiQB
zUt_(}4^6eR`8$8J-lb<TtsAP@C(JNfmU5~`=FEA2pF4Ntc1u^w$Fo!zJa%hmv$5(@
z{iW4WdU{n>_SLIeD_3nbIn_1g%IT7;rG?DT-}d(Uuk|_a?{kK)?_7_#Zl9j+bTge?
z-(H_GqdbYonKSZK&TQY8uy^}Rqf5`8T}oLk?JXIU?l<|o`uyqBuRcG(z`(%xLgdP=
z`BBpX1H&R@W#>w0PPsgJ%F8tow`NPtc#|BwUT4kv;A1oMYYtZD8_C9+n#Nd$nN2fV
z?AR<l&);5Le1+BDZ-rm}W}Vr+<?8-7HasFTmitY=G38>e^2Zz3uG>GqbFROx+c3SY
z<jHB*ZA<@Dw%xoFVI*p5J?rqI$-(;=w>|C6T5~#+&(-W~K#BW>+j4ViO-i>%&&dpb
zRjj_S?zDRPzo^uwQQ2}er>E*Jwb;0RTZ~4$>6H{Q9^b%(y-mp}b!yr3*q+9OZ8n>K
zHE8>bC2`Ad5zK(h52Qmk{n+gL^|(o{T=urN5~tade}Dg$-)LLJdh~rp_qT5;OXYR;
zg~x9`A<5W0ZKbrVUWCDw%IModug*m73C?|7uYdoX!r!;n`cpps+wpSRI?dZ<IjNg0
znGNip?z?Z>x9j2Jcdx~TD}+Ml=k9CSU*7-3AdOSse09(1>-$CD%RRc5mAPed$WQAT
znawqO@7KB(>!t2`dhztpzmC_BUSpHE@r!3oZ1<<M)cVlL2c7wrJvY6TYE`pha?rsS
zlYMgzhivX#dF}O-9kM#YGo+2)E8W$;(i@QOw_i48oAS{;{~1>J9nk+#U;WnJ>7B;b
z`e%Qdx77S+nDY0QMxpwgn*Ay5H*Kz;ut|R${o{VtC+>GoR|v4x|3mGON;urOapl*Q
zGhe=(Iq>4mi-Scs9zD8QoOJBUjVs5K)gHFaoZ0rs=XCWm?%S1<b@lRc)x0)t**I0L
z<fYB0%Ab|BaW&BVYyIcj;xBst87}==khe|hKf{ulTX;>M`Tsq4s_x>3$DUK_|Ax;F
zm&ku_|0?HVNbi4!?Rz#Pvu~e$TRV1J*H-DRS2}f@<hE?n|9tBD+2SX=vYfKcZM}Z=
z&K%R9YrlOAx>|TspZ|l~)R)_jObsy)(TyxSeldE&k{h89GqNv!J}Y<S%GH3WVS#~R
z;g>E2U4IEq)vKqAhFlHTUJOmu?SW4FT?6^RsoG-e)zXW}j?L0d=Ir9)OKhQDz(X+}
zkl$DQd)<C%`<F8BKRn>SZ~6CneDC}(ee4Y2#)EX-)pJdIBg)-b7U%5Sk~ROz8r4aA
zW*?ecxAS#YdDVyKVj^tzXOI6mvg&s1?b)IEtLoRTNT_*!qUI0(tTj`)avH91SjQfk
zcjE7beOVjYH_q-~Be7P!w=QJXHW^LlshdBC99!?&8!jT1&GLNd+Uyl;<0l55t+Wa@
zDPNr}QD%BIv_`9Vam<+=i(XXw{9PKWWjgtI<YuO84tM^{4*YWJ`@HK`Mnxe>r=D+#
z?@l<Hu(s9k`i*t_)B8U$v+MKB-V;4<T^ZZ_OFIJtmg}9nC(`!P_n?Wb)dky=uNb}?
z%Vs{?`-JDw!N5=dd=qx_@j5BlpTE5M-sM`?Q@)e0f1GCiDZ;bo50_opo!!sGMP**E
z`g^x|nrXVJxK&uS_f~noHQ~l)FYoBTdsg-!e!l#4sm&^vysngVi%ktZAueOM@pNzM
z*?n0F{@Z`cYft$k=Js{H*3BTp%f=rci0qlPV$a3UEAMoV{MgK{6qs~Jv)AyS{4uY;
z=Z_U$)Afxx{-S&DyGIARo^RAxdF_h+mGUy~>gd~c_v7WCztef6^Ry;%bIRvWfAK^r
zLW)52vexZgXTM;vTfMOTKDmQJtf-9w-v0=L2(V-dtFK?VnpSfA+VYj%+Mni5H?MW|
ztvu#yT5cUGw0ara`RjbzCTDCm?hX5z{dU?5k8M3yE~FKN$44w(d2UOume*5z9lPUx
z4trnSUv!%BKg08XI)Cq$t>A0|St};>?exm6s$VQGN|lRFW3{gA{<(FNY_zFqFK^G%
z@Vm3l>U*rYxZsvsxBA@Li>keqNBy36cVCS;wz}Hqu)kWW>k`R1AD`-ezEu7tvgGoA
zh6B9jfA1Eako*TSn;2Yl^gZ{X-pJs^U)q<xX8lnrb$#;>Z_~Z*JEtbg9lII!P59i-
z!|Q_|Hisrl6|+lktMl73S?p0NZ~4Not+UES*PHIK+GRa&xxd|wWvhOzI@9%K)4msz
zUhFSg_Gr#c-<V^oZmc>Ut2Vbab!OEglT+PKf8Xkyto1H)t>US5Th>ig<9TuC<IbNI
zdmWhy!m=}Lm(|K|vt51R(T3(3rhfO5nHL*}*v6j>SP>e%Cgx4Tt=fruGj^7G8!cI&
zztXcm`N*{8ZQ&aytyL|3^Qu19<W$$u2<;sQ|8mWn7XSF!e}?}Iv#MVIyXr7QLGwGh
zp!*lCs>u5O+5Z{-WxihjJCrG*^;$_cmuYs~^jB&-_eSX7woZ^uI1mwd?4n<2&(mnN
zHQ`+$7b9}S(qlpcZl8&tneS@kYO>nPDqB+Y&cvO|L&H74emeZ{=wGWB2QM=_&U==s
z8+JvezW-BB|Ma&%MZ4bqXE3l`S))}}dNbZ{_U7n1mxN=oKWk5Rb?$z7wQ2vX%V!^~
z$$MY0T5E^Gn?J<F8aZkhKmPeUf5Wx4iEF^^ndu+nq>pC*fwX7-b-n6(?`{Ka&z!El
z#(leTD!4s!?W~_~*p*XVQ?BbwH!Ustcc6g%efYiUyR3UtUOsAhK0!2O*2cf<PULu<
ztrxW|QVU<cvHGKJ?efh#+pVwNu6&@oW3jGj+>%I_buLeKY&^QoW=^i}sx3wV&JnWi
zdvEP^G)sN^*WXvN?%A%>+UKT({MBBaE&R>$s^GkNOY4lqTu)xN&+fk*(<uH#@8*=s
zuh;zSb{rF$Q~7e=neIvJCSNPEzkOCL`$o*~O}(M-qb27Qv(=TYjPF_=_ITE(=w%^U
zYeSopvkO_%_O`xy^$XlBy?xqa%f)Af*WEYn+&cHL^hLE}>H6l!4=`WQ+-9n1s<T&V
zvw`Yv)eXUB>J}^S+*oI-xm7RX*p9fg$i(H7`R=tgb={M(_R5~Mb7D@`o1|lN+5B1N
z@#)3;&u7T54)}ZWZ$c^CG4VxwR}6Bh{H0kxn;x^gec$EI-<XrhWvh>@KPr{ymAIvS
zLy7j6m<g|Lz6iNkY}B<TZPmMI$tjl?8(JJlS$5ycZuP9kS69a*J=>$H4eA<aEINOK
z?b|)w?t6UI8@ttK+19r7NnE`b+J9?z>Do5kQ=wl&ue_h~^;%(SnAQ{4nAZNFq@MiW
zxBd6(S7l#~j|q8f`GiBlJ7*KqyY0_vauy%;&-9;=XA~}b$c$x2(&FT~kdE=AWoOU5
z_f^xrbx7UJby?(-9cv2HJqx&qD~A~v7=%>62X8Yq{Ce#6tTM4Q?pH0UfAfM}SMS`p
zamzaI(3xv?biPVDHAnJN<zqelQ)MNEg_C{#kMixFyK>{q_G#v;?)|!yvg)l6>v#9M
zxwX&!7}q_Y^^2{+fq{X6-(lH}D=k+xed#)LVA`8$2YqhLd9=JZX~&giS9T^#J*=FW
z+BVO}x=M^oyQ1K&=DMtBt{Ycx?3OB#v=OSX*jZ~=!&ndQ5fYJOm>l-Kcz=04s3&>n
zbG+AI;gs*oUp@|syOaIy-Fn6K2aa0q=f9rlzFgF`T4I~g=69OAb9S^RPM?vm#x?qR
zLdoHr%+nF;LJy~IUM#havv}E~RqMiziPr3ScvJvnIotJL3Ym|?-j}*v{^Rb~w(sAh
zjWdLe<H}Z7hHuO|R#U;vYaF~c<=3CdY+8p?-51t=TeQAg=;G1BwSChrPuL{oCC$4w
zChfk@IfE@HK*kY)89SseFS_;kN_<Sr*ZK@$o7M7LYqYJhXPt}KdcNZ5yIndjPFs45
zZT4(VezcP@e8;&j;Ww%t+_-$_@{M~-cLkaze_M9c`p$j^x9O|oonk+~EV}ZxaKgvg
zY#~=qrEy-IyzQ|}VAb1=a}TXwchqTP)pPZ*;2RsGUq5EuyZh0t@}=q8MOUuJJ<)e&
z`lg+8YD&o+-)m8c`X7?mf?mE|u=D7<@0w|Q<*x7?oE&AHJ~<_3=~|0;1>22fUR7Uj
z<^8&IV`|>P{9EF$v{tOPZf)tGuzbz>HQgT;I`tkrB5;g#U%kqEZ~d#w7szkix9dLx
zpZk-!FH?VQ`NI6o{CfPu6M{!CuVub7ZvlU?;k9|$_Z-gYh|23fxyHTW!4)OtKYy)l
z-<n^ES#s01UNX5}wPnWn1B;*Mp5B*O7<g{i`nxwQlf#Nb^A{d>?lHY2&RZN6RkZ%8
z?}by{A09pL`6woDzi!v6-Ph0a-F|y$M)2Eh?R@vz`7cHvZZ^7hEk$bDnp0PU<