kleine Verbesserungen

projectmood/face_detect.py

@@ -6,9 +6,9 @@ faceDet_three = cv2.CascadeClassifier('resources/haarcascade/haarcascade_frontal
 faceDet_four = cv2.CascadeClassifier('resources/haarcascade/haarcascade_frontalface_alt_tree.xml')
 
 
-def extract_faces(image):
+def extract_faces(image, resize):
     faces = find_faces(image)
-    normalized_faces = [preprocess_face(face) for face in faces]
+    normalized_faces = [preprocess_face(face, resize) for face in faces]
     return normalized_faces
 
 def find_faces(image):
@@ -17,9 +17,9 @@ def find_faces(image):
     return cutted_faces
 
 
-def preprocess_face(face):
+def preprocess_face(face, resize):
     face = cv2.cvtColor(face, cv2.COLOR_BGR2GRAY)
-    face = cv2.resize(face, (250, 250))
+    face = cv2.resize(face, (resize, resize))
     clahe = cv2.createCLAHE(clipLimit=2.0, tileGridSize=(8, 8))
     face = clahe.apply(face)
     return face

projectmood/image_sort.py

@@ -14,7 +14,7 @@ emotions = ["neutral", "anger", "contempt", "disgust", "fear", "happy", "sadness
 """
 Liste mit den Dateiordnern aller Teilnehmer 
 """
-participants = glob.glob("img_data/source_emotion/*")
+participants = glob.glob("resources/img_data/source_emotion/*")
 
 for x in participants:
 
@@ -27,7 +27,7 @@ for x in participants:
             """
             Sessionnummer
             """
-            current_session = files[31:-30]
+            current_session = files[39:-30]
 
             """
             Öffnet die zur aktuellen Emotion korrelierende .txt Datei
@@ -35,11 +35,11 @@ for x in participants:
             file = open(files, 'r')
 
             """
-            In der Datei steht die aktuell betrachtete Emotion, kodiert als folat-Wert 
+            In der Datei steht die aktuell betrachtete Emotion, kodiert als float-Wert 
             """
             emotion = int(float(file.readline()))
 
-            source_emotions = glob.glob("img_data/source_images/%s/%s/*.png" % (number, current_session))
+            source_emotions = glob.glob("resources/img_data/source_images/%s/%s/*.png" % (number, current_session))
             source_emotions.sort()
 
             """
@@ -58,8 +58,8 @@ for x in participants:
             Für den neutralen Ausdruck 
             Für die Emotion 
             """
-            dest_neut = "img_data/sorted_set/neutral/%s" % sourcefile_neutral[36:]
-            dest_emot = "img_data/sorted_set/%s/%s" % (emotions[emotion], sourcefile_emotion[36:])
+            dest_neut = "resources/img_data/sorted_set/neutral/%s" % sourcefile_neutral[44:]
+            dest_emot = "resources/img_data/sorted_set/%s/%s" % (emotions[emotion], sourcefile_emotion[44:])
 
             """Kopiert Dateien"""
             copyfile(sourcefile_neutral, dest_neut)

projectmood/process_model.py

@@ -28,9 +28,9 @@ Argument Parser erlaubt Parameter für die Verarbeitung anzugeben.
 parser = argparse.ArgumentParser(description='Process Model Application')
 parser.add_argument('--dataset', action='store', dest='dataset', default='resources/img_data/dataset/', help='path to dataset')
 parser.add_argument('-i', action='store', dest='iterations', type=int, default=30, help='number of iterations')
-parser.add_argument('-e', action='append', dest='emotions', default=['happy', 'neutral', 'surprise'], help='declare emotions that should be processed')
+parser.add_argument('-e', action='append', dest='emotions', default=['happy', 'neutral', 'sadness', 'surprise'], help='declare emotions that should be processed')
 parser.add_argument('-p', action='append', dest='properties', help='pre-processing steps for logging')
-parser.add_argument('--test', action='store_true', help='prevent writing new model to file system')
+parser.add_argument('--test', action='store_true', help='prevent writing new model to classifier')
 parser.add_argument('--csv', action='store_true', help='activate csv processing')
 parser.add_argument('--email', action='store_true', help='activate email notifications')
 arguments = parser.parse_args()
@@ -78,9 +78,9 @@ def image_preprocessing(image):
     """
     img = cv2.imread(image)  # open image
     gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)  # convert to grayscale
-    clahe = cv2.createCLAHE(clipLimit=2.0, tileGridSize=(8, 8))
-    face = clahe.apply(gray)
-    return face
+    clahe = cv2.createCLAHE(2.0, (8,8))
+    norm = clahe.apply(gray)
+    return norm
 
 
 def make_sets():
@@ -146,7 +146,7 @@ if csv:
 logging.info("Fisherface training finished - {}% average\n".format(np.mean(metascore)))
 
 if not test:
-    fishface.write('img_data/models/detection_model.xml')
+    fishface.write('resources/models/detection_model.xml')
 
 if email:
     sendMail('Fisherface training finished')

projectmood/sorted_set_facedetector.py

@@ -24,7 +24,7 @@ parser = argparse.ArgumentParser(description='Sorted Set Face Creator Applicatio
 parser.add_argument('--source', action='store', dest='img_source', default='resources/img_data/sorted_set/',
                     help='path to image source')
 parser.add_argument('--dataset', action='store', dest='dataset', default='resources/img_data/dataset/', help='path to dataset')
-parser.add_argument('--r', action='store', dest='resize', default=150, type=int, help='resize factor')
+parser.add_argument('-r', action='store', dest='resize', default=150, type=int, help='resize factor')
 parser.add_argument('-e', action='append', dest='emotions', default=['happy', 'neutral', 'surprise'], help='declare emotions that should be processed')
 parser.add_argument('-c', action='store', dest='scaleFactor', default=1.1, type=float,
                     help='scale factor - haar')
@@ -79,7 +79,7 @@ def detect_faces(emotion):
         facefeatures = locate_faces(gray, scaleFactor, minNeighbors, (minSize, minSize))
 
         if facefeatures is '':
-            logging.info('No face detected ' + f)
+            logging.info('No face detected in ' + f)
             undetected.append(f)
             if len(undetected) % 200 == 0 and email:
                 sendMail('Already ' + str(len(undetected)) + ' not detected faces', filepath=logfile)
@@ -94,7 +94,7 @@ def detect_faces(emotion):
                     if not success:
                         logging.error('Problem while writing file ' + f + ' occurred...' )
                         if email:
-                            sendMail('Problem while writing file', body=f + ' to ' + datasetPath + '{}/{}.jpg'.format(emotion, fileNumber))
+                            sendMail('Problem while writing file', body=f + ' to ' + dataset_path + '{}/{}.jpg'.format(emotion, fileNumber))
                 except:
                     logging.error('Some error with ' + f)
                     if email:

projectmood/webcam.py

@@ -2,6 +2,7 @@
 Diese Klasse ist das Main-Modul. Es lädt das Modell aus models, zeigt ein Webcam-Bild,
 erkennt das Gesicht und seine Emotionen und zeichnet ein Emoticon in das Bild.
 """
+import argparse
 
 import cv2
 
@@ -11,6 +12,12 @@ from face_detect import extract_faces
 from image_commons import nparray_as_image, draw_with_alpha
 import numpy as np
 
+parser = argparse.ArgumentParser(description='ProjectMood Camplication')
+parser.add_argument('-b', action='store', dest='buffer', default=10, type=int, help='size of ringbuffer')
+parser.add_argument('-e', action='append', dest='emotions', default=['happy', 'neutral', 'sadness', 'surprise'], help='declare emotions that should be detected')
+parser.add_argument('-r', action='append', dest='resize', default=150, help='resize factor')
+parser.add_argument('--model', action='store', dest='model', default='resources/models/detection_model.xml', help='path to model')
+arguments = parser.parse_args()
 
 def _load_emoticons(emotions):
     """
@@ -37,14 +44,14 @@ def show_webcam_and_run(model, emoticons, window_size=(600, 600), window_name='P
     vc = WebcamVideoStream().start()
 
     """
-    Der RingBuffer speichert die letzten 7 Predictions
+    Der RingBuffer speichert die letzten x Predictions
     """
-    buffer = RingBuffer(7)
+    buffer = RingBuffer(arguments.buffer)
 
     frame = vc.read()
 
     while True:
-        for normalized_face in extract_faces(frame):
+        for normalized_face in extract_faces(frame, arguments.resize):
             prediction = model.predict(normalized_face)  # do prediction
 
             """
@@ -56,6 +63,7 @@ def show_webcam_and_run(model, emoticons, window_size=(600, 600), window_name='P
             Holt die Einträge als Array
             """
             predictions = buffer.get()
+            print(predictions)
 
             """
             Kein Eintrag im RingBuffer ist None 
@@ -84,11 +92,11 @@ def show_webcam_and_run(model, emoticons, window_size=(600, 600), window_name='P
 
 
 if __name__ == '__main__':
-    emotions = ['happy', 'neutral', 'surprise']
+    emotions = arguments.emotions
     emoticons = _load_emoticons(emotions)
 
     fisher_face = cv2.face.FisherFaceRecognizer_create()
 
     """Lädt das trainierte Model"""
-    fisher_face.read('resources/models/detection_model.xml')
+    fisher_face.read(arguments.model)
     show_webcam_and_run(fisher_face, emoticons)