Committing the logic folder

LecturerApp/logic/classroom_activity.py

+import tensorflow as tf
+import tensorflow.keras
+from PIL import Image, ImageOps
+import numpy as np
+import cv2
+import os
+
+def activity_recognition(video_name):
+    BASE_DIR = os.path.dirname(os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
+    CLASSIFIER_DIR = os.path.join(BASE_DIR, "LecturerApp\\models")
+    VIDEO_PATH = os.path.join(BASE_DIR, "LecturerStatic\\LecturerApp\\videos\\{}".format(video_name))
+
+    print('video name: ', video_name)
+
+    detector_path = os.path.join(CLASSIFIER_DIR, "keras_model.h5")
+
+    # Disable scientific notation for clarity
+    np.set_printoptions(suppress=True)
+
+    class_labels = ['Seated Teaching', 'Teaching by Standing', 'Teaching by Walking']
+
+
+    # img = cv2.imread('test2.jpg')
+
+    # Load the model
+    model = tensorflow.keras.models.load_model(detector_path)
+    model.compile(optimizer='adam',
+                  loss=tf.keras.losses.SparseCategoricalCrossentropy(from_logits=True),
+                  metrics=['accuracy'])
+
+
+
+
+
+
+    # Create the array of the right shape to feed into the keras model
+    # The 'length' or number of images you can put into the array is
+    # determined by the first position in the shape tuple, in this case 1.
+    data = np.ndarray(shape=(1, 224, 224, 3), dtype=np.float32)
+    size = (224, 224)
+    # "C://Users//DELL//Downloads//Classroom_Video.mp4"
+    video1 = 'E:\\Studies\\SLIIT\\4th Year\\Python Projects\\classroom activity models\\videos\\{}'.format(video_name)
+
+    video = cv2.VideoCapture(VIDEO_PATH)
+
+    # additional
+    # getting the frames per second (fps)
+    # fps = video.get(cv2.CAP_PROP_FPS)
+
+    # getting the number of frames using CAP_PROP_FRAME_COUNT method
+    no_of_frames = video.get(cv2.CAP_PROP_FRAME_COUNT)
+    frame_count = 0
+    seated_count = 0
+    standing_count = 0
+    walking_count = 0
+
+    # while loop is conditioned like this to avoid the termination of the loop with an exception
+    while (frame_count < no_of_frames):
+
+        cap, frame = video.read()
+        # gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
+        img = cv2.resize(frame, size)
+        # image_array = np.asarray(img)
+        normalized_image_array = (img.astype(np.float32) / 127.0) - 1
+
+        # Load the image into the array
+        data[0] = normalized_image_array
+
+        # run the inference
+        prediction = model.predict(data)
+        label = class_labels[prediction.argmax()]
+
+        if (label == class_labels[0]):
+            seated_count += 1
+        elif (label == class_labels[1]):
+            standing_count += 1
+        elif (label == class_labels[2]):
+            walking_count += 1
+        # text = label + format(prediction[prediction.argmax()])
+
+        # if label:
+        cv2.putText(frame, label, (20, 60), cv2.FONT_HERSHEY_SIMPLEX, 1, (0, 255, 0), 3)
+        # else:
+        #     cv2.putText(frame, 'Unknown', (20, 60), cv2.FONT_HERSHEY_SIMPLEX, 2, (0, 255, 0), 3)
+
+        frame_count += 1
+        # cv2.imshow('Activity Recognition', frame)
+        # if cv2.waitKey(1) & 0xFF == ord('q'):
+        #     break
+
+
+
+    cv2.destroyAllWindows()
+    print("No of frames: ", frame_count)
+
+    # calculating the percentages
+    sit_perct = (seated_count / no_of_frames) * 100
+    stand_perct = (standing_count / no_of_frames) * 100
+
+    walk_perct = (walking_count / no_of_frames) * 100
+
+    return {
+        "sitting_perct": sit_perct,
+        "standing_perct": stand_perct,
+        "walking_perct": walk_perct
+    }
+
+
+