run_code.py

from tensorflow.keras.layers import Input

from src.yolo3.model import *
from src.yolo3.detect import *

from src.utils.image import *
from src.utils.datagen import *
from src.utils.fixes import *

fix_tf_gpu()
def prepare_model(approach):
    '''
    Prepare the YOLO model
    '''
    global input_shape, class_names, anchor_boxes, num_classes, num_anchors, model

    # shape (height, width) of the imput image
    input_shape = (416, 416)

    # class names
    if approach == 1:
        class_names = ['H', 'V', 'W']

    elif approach == 2:
        class_names  = ['W','WH','WV','WHV']

    elif approach == 3:
        class_names  = ['W']

    else:
        raise NotImplementedError('Approach should be 1, 2, or 3')

    # anchor boxes
    if approach == 1:
        anchor_boxes = np.array(
            [
            np.array([[ 76,  59], [ 84, 136], [188, 225]]) /32, # output-1 anchor boxes
            np.array([[ 25,  15], [ 46,  29], [ 27,  56]]) /16, # output-2 anchor boxes
            np.array([[ 5,    3], [ 10,   8], [ 12,  26]]) /8   # output-3 anchor boxes
            ],
            dtype='float64'
        )
    else:
        anchor_boxes = np.array(
            [
            np.array([[ 73, 158], [128, 209], [224, 246]]) /32, # output-1 anchor boxes
            np.array([[ 32,  50], [ 40, 104], [ 76,  73]]) /16, # output-2 anchor boxes
            np.array([[ 6,   11], [ 11,  23], [ 19,  36]]) /8   # output-3 anchor boxes
            ],
            dtype='float64'
        )

    # number of classes and number of anchors
    num_classes = len(class_names)
    num_anchors = anchor_boxes.shape[0] * anchor_boxes.shape[1]

    # input and output
    input_tensor = Input( shape=(input_shape[0], input_shape[1], 3) ) # input
    num_out_filters = ( num_anchors//3 ) * ( 5 + num_classes )        # output

    # build the model
    model = yolo_body(input_tensor, num_out_filters)

    # load weights
    weight_path = f'model-data/weights/pictor-ppe-v302-a{approach}-yolo-v3-weights.h5'
    model.load_weights( weight_path )


def get_detection(img):
    # save a copy of the img
    act_img = img.copy()

    # shape of the image
    ih, iw = act_img.shape[:2]

    # preprocess the image
    img = letterbox_image(img, input_shape)
    img = np.expand_dims(img, 0)
    image_data = np.array(img) / 255.

    # raw prediction from yolo model
    prediction = model.predict(image_data)

    # process the raw prediction to get the bounding boxes
    boxes = detection(
        prediction,
        anchor_boxes,
        num_classes,
        image_shape=(ih, iw),
        input_shape=(416, 416),
        max_boxes=10,
        score_threshold=0.3,
        iou_threshold=0.45,
        classes_can_overlap=False)

    # convert tensor to numpy
    boxes = boxes[0].numpy()

    # draw the detection on the actual image
    return (draw_detection(act_img, boxes, class_names), boxes)



def run (image_in, approach):
    prepare_model(approach=approach)

#    input_shape = (416, 416)
    img = letterbox_image(image_in, input_shape)

    # get the detection on the image
    img, all_classes = get_detection(img)

    #print (all_classes)
    WHV = 0
    WV = 0
    WH = 0
    W = 0
    H = 0
    V = 0
    for i in all_classes:
        if class_names[int(i[-1])] == "WHV":
            WHV += 1
            W += 1
        elif class_names[int(i[-1])] == "WH":
            WH += 1
            W += 1
        elif class_names[int(i[-1])] == "H":
            H += 1
        elif class_names[int(i[-1])] == "V":
            V += 1
        elif class_names[int(i[-1])] == "WV":
            WV += 1
            W += 1
        elif class_names[int(i[-1])] == "W":
            W += 1

    #Outputs to display the number of each classes in an interpretable format
    texts = ""
    texts = texts + "Total workers: " + str(W) + "\n"
    if approach != 3:
        if approach == 1:
            texts = texts + "Number of Helmets: " + str(H) + "\n"
            texts = texts + "Number of Vests: " + str(V) + "\n"

        elif approach == 2:

            texts = texts + "Workers wearing helmet and vest: " + str(WHV) + "\n"
            texts = texts + "Workers wearing only vest: " + str(WV) + "\n"
            texts = texts + "Workers wearing only helmet: " + str(WH) + "\n"

            if (W > WHV) and (WHV != 0):
                texts = texts + "Workers not wearing helmet and vest: " + str(W - WHV) + "\n"

            if (W > WH) and (WH != 0):
                texts = texts + "Workers not wearing helmet and vest: " + str(W - WH) + "\n"

            if (W > WV) and (WV != 0):
                texts = texts + "Workers not wearing helmet and vest: " + str(W - WV) + "\n"

    return {'img': img[:, :, ::-1], 'text': texts}