Exercise-App/main.py at main · OhRai/Exercise-App · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
import cv2
import numpy as np
import tensorflow as tf
from PIL import ImageFont, ImageDraw, Image

# Load Model
from utils.models import model
from utils.keypoints import draw_keypoints, draw_edges, edges

# Initialize TensorFlow Lite Interpreter
model_type, input_size = model('thunder')
interpreter = tf.lite.Interpreter(model_path=model_type)
interpreter.allocate_tensors()

# Exercise Class Initializer
from exercises.push_up import PushUpCounter
from exercises.bicep_curl import CurlCounter
from exercises.squat import SquatCounter
from exercises.deadlift import DeadliftCounter

push_up = PushUpCounter()
curl_counter = CurlCounter()
squat_counter = SquatCounter()
deadlift_counter = DeadliftCounter()

# Load Interface and Video Capture
interface = cv2.imread('app_ui/main.png')
cap = cv2.VideoCapture('data/video7.mp4')
clean_interface = interface.copy()

# Initialize UI Position and Exercise Variables
push_pos = [(720, 308), (720+240, 308+65)]
bicep_pos = [(1000, 308), (1000+240, 308+65)]
squat_pos = [(720, 411), (720+240, 411+65)]
deadlift_pos = [(1000, 411), (1000+240, 411+65)]

font = ImageFont.truetype('app_ui/roboto.ttf', 36)
total_reps = 0
exercise = 'Push-up'

# Button Click Event Handler
def button_click(event, x, y, flags, param):
    global exercise, total_reps, push_up, curl_counter, squat_counter, deadlift_counter
    if event == cv2.EVENT_LBUTTONDOWN:
        if push_pos[0][0] <= x <= push_pos[1][0] and push_pos[0][1] <= y <= push_pos[1][1]:
            exercise = 'Push-up'
            push_up = PushUpCounter()
        if bicep_pos[0][0] <= x <= bicep_pos[1][0] and bicep_pos[0][1] <= y <= bicep_pos[1][1]:
            exercise = 'Bicep Curl'
            curl_counter = CurlCounter()
        if squat_pos[0][0] <= x <= squat_pos[1][0] and squat_pos[0][1] <= y <= squat_pos[1][1]:
            exercise = 'Squat'
            squat_counter = SquatCounter()
        if deadlift_pos[0][0] <= x <= deadlift_pos[1][0] and deadlift_pos[0][1] <= y <= deadlift_pos[1][1]:
            exercise = 'Deadlift'
            deadlift_counter = DeadliftCounter()

# Set up UI Window and Mouse Click Event
cv2.namedWindow('Exercise App')
cv2.setMouseCallback('Exercise App', button_click)

# Video Loop
while cap.isOpened():
    # Setup Background and Camera
    interface = clean_interface.copy()
    ret, frame = cap.read()

    if not ret:
        break

    # Ensure Frame Aspect Ratio
    width = int(cap.get(cv2.CAP_PROP_FRAME_WIDTH))
    height = int(cap.get(cv2.CAP_PROP_FRAME_HEIGHT))

    if (width > height):
        border_size = (width - height) // 2
        frame = cv2.copyMakeBorder(frame, border_size, border_size, 0, 0, cv2.BORDER_CONSTANT, value=[0, 0, 0])

    else:
        border_size = (height - width) // 2
        frame = cv2.copyMakeBorder(frame, 0, 0, border_size, border_size, cv2.BORDER_CONSTANT, value=[0, 0, 0])

    frame = cv2.resize(frame, (640, 640))

    # Model Camera
    img = frame.copy()
    img = tf.image.resize_with_pad(np.expand_dims(img, axis=0), 256, 256)
    input_image = tf.cast(img, dtype=tf.float32)

    input_details = interpreter.get_input_details()
    output_details = interpreter.get_output_details()

    interpreter.set_tensor(input_details[0]['index'], np.array(input_image))
    interpreter.invoke()
    keypoints_with_scores = interpreter.get_tensor(output_details[0]['index'])

    # Process Functions and Exercises
    draw_keypoints(frame, keypoints_with_scores, 0.2)
    draw_edges(frame, keypoints_with_scores, edges, 0.2)

    if exercise == 'Push-up':
        total_reps = push_up.count_pushup(frame, keypoints_with_scores, 0.2)
    elif exercise == 'Bicep Curl':
        total_reps = curl_counter.count_curl(frame, keypoints_with_scores, 0.2)
    elif exercise == 'Squat':
        total_reps = squat_counter.count_squat(frame, keypoints_with_scores, 0.2)
    elif exercise == 'Deadlift':
        total_reps = deadlift_counter.count_deadlift(frame, keypoints_with_scores, 0.2)

    # Update Frame and Text
    img_pil = Image.fromarray(interface)
    draw = ImageDraw.Draw(img_pil)

    draw.text((1040, 62), exercise, font=font, fill=(0, 0, 0), antialias=True)
    draw.text((1040, 113), str(total_reps), font=font, fill=(0, 0, 0), antialias=True)

    interface = np.array(img_pil)
    interface[40:640+40, 40:640+40] = frame

    # Show Frame
    cv2.imshow('Exercise App', interface)

    if cv2.waitKey(1) & 0xFF == ord('q'):
        break

cap.release()
cv2.destroyAllWindows()