norep-app/mediapipe_stream.py at main · andersgom/norep-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
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
# PACKAGES

import cv2
import mediapipe as mp
import numpy as np
import pandas as pd
mp_drawing = mp.solutions.drawing_utils
mp_pose = mp.solutions.pose
import pickle
from helper_funcs import calculate_angle


with open('repcounter.p', 'rb') as file:
    model = pickle.load(file)


# NOREP APP - MEDIAPIPE STREAM

cap = cv2.VideoCapture(0)

# Counter variables
counter = 0
grip = None
stance = None
stage = None

## Setup mediapipe instance

with mp_pose.Pose(min_detection_confidence=0.5, min_tracking_confidence=0.5) as pose:
    while cap.isOpened():
        ret, frame = cap.read()

        # Recolor image to RGB
        image = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)
        image.flags.writeable = False

        # Make detection
        results = pose.process(image)

        # Recolor back to BGR
        image.flags.writeable = True
        image = cv2.cvtColor(image, cv2.COLOR_RGB2BGR)

        # Extract landmarks
        try:
            landmarks = results.pose_landmarks.landmark

            # Get coordinates
            # Grip
            l_shoulder = [landmarks[mp_pose.PoseLandmark.LEFT_SHOULDER.value].x,landmarks[mp_pose.PoseLandmark.LEFT_SHOULDER.value].y]
            r_shoulder = [landmarks[mp_pose.PoseLandmark.RIGHT_SHOULDER.value].x,landmarks[mp_pose.PoseLandmark.RIGHT_SHOULDER.value].y]
            l_elbow = [landmarks[mp_pose.PoseLandmark.LEFT_ELBOW.value].x,landmarks[mp_pose.PoseLandmark.LEFT_ELBOW.value].y]
            r_elbow = [landmarks[mp_pose.PoseLandmark.RIGHT_ELBOW.value].x,landmarks[mp_pose.PoseLandmark.RIGHT_ELBOW.value].y]
            # Stance
            l_hip = [landmarks[mp_pose.PoseLandmark.LEFT_HIP.value].x,landmarks[mp_pose.PoseLandmark.LEFT_HIP.value].y]
            r_hip = [landmarks[mp_pose.PoseLandmark.RIGHT_HIP.value].x,landmarks[mp_pose.PoseLandmark.RIGHT_HIP.value].y]
            l_ankle = [landmarks[mp_pose.PoseLandmark.LEFT_ANKLE.value].x,landmarks[mp_pose.PoseLandmark.LEFT_ANKLE.value].y]
            r_ankle = [landmarks[mp_pose.PoseLandmark.RIGHT_ANKLE.value].x,landmarks[mp_pose.PoseLandmark.RIGHT_ANKLE.value].y]


            # Calculate angles
            l_grip = calculate_angle(r_shoulder, l_shoulder, l_elbow)
            r_grip = calculate_angle(l_shoulder, r_shoulder, r_elbow)
            l_stance = calculate_angle(r_hip, l_hip, l_ankle)
            r_stance = calculate_angle(l_hip, r_hip, r_ankle)

            # Visualize angle  -  FOR DEBUGGING PURPOSES
            #cv2.putText(image, str(l_grip),
            #               tuple(np.multiply(l_shoulder, [640, 480]).astype(int)),
            #               cv2.FONT_HERSHEY_SIMPLEX, 0.5, (255, 255, 255), 2, cv2.LINE_AA
            #                    )
            #cv2.putText(image, str(r_grip),
            #               tuple(np.multiply(r_shoulder, [640, 480]).astype(int)),
            #               cv2.FONT_HERSHEY_SIMPLEX, 0.5, (255, 255, 255), 2, cv2.LINE_AA
            #                    )
            #cv2.putText(image, str(l_stance),
            #               tuple(np.multiply(l_ankle, [640, 480]).astype(int)),
            #               cv2.FONT_HERSHEY_SIMPLEX, 0.5, (255, 0, 255), 2, cv2.LINE_AA
            #                    )
            #cv2.putText(image, str(r_stance),
            #               tuple(np.multiply(r_ankle, [640, 480]).astype(int)),
            #               cv2.FONT_HERSHEY_SIMPLEX, 0.5, (255, 0, 255), 2, cv2.LINE_AA
            #                    )


            # Grip logic
            if (r_grip>90)|(l_grip>90)&(r_grip<120)|(l_grip<120):
                grip = 'Grip: Good!'
                posturebox = cv2.rectangle(image, (0,150), (225,73), (200,200,16), -1)
            if (r_grip>120)|(l_grip>120):
                grip = 'Grip: Too wide'
                posturebox = cv2.rectangle(image, (0,150), (225,73), (0,145,218), -1)
            if (r_grip<90)|(l_grip<90):
                grip = 'Grip: Too narrow'
                posturebox = cv2.rectangle(image, (0,150), (225,73), (0,145,218), -1)

            # Stance logic
            if (r_stance>88)|(l_stance>88)&(r_stance<98)|(l_stance<98):
                stance = 'Stance: Good!'
                posturebox = cv2.rectangle(image, (0,150), (225,73), (200,200,16), -1)
            if (r_stance>98)|(l_stance>98):
                stance = 'Stance: Too wide'
                posturebox = cv2.rectangle(image, (0,150), (225,73), (0,145,218), -1)
            if (r_stance<88)|(l_stance<88):
                stance = 'Stance: Too narrow'
                posturebox = cv2.rectangle(image, (0,150), (225,73), (0,145,218), -1)

            # Model implementation
            poses = results.pose_landmarks.landmark
            pose_row = np.array([[landmark.x, landmark.y, landmark.z] for landmark in poses]).flatten()
            frame_height, frame_width = frame.shape[:2]
            pose_row = pose_row * np.array([frame_width, frame_height, frame_width])[:,None]
            X = pd.DataFrame([pose_row[0]])
            body_language_class = model.predict(X)[0]
            body_language_prob = model.predict_proba(X)[0]

            # Rep counter logic
            if body_language_class == 0:
                stage = 'Down'
            if (body_language_class == 1)&(stage=='Down'):
                stage = 'Up'
                counter +=1

        except:
            pass


        # Setup status box
        cv2.rectangle(image, (0,0), (225,73), (87,122,59), -1)
        postureboxlogic = posturebox
        postureboxlogic

        # Rep data
        cv2.putText(image, 'REPS', (25,15),
                    cv2.FONT_HERSHEY_SIMPLEX, 0.5, (0,0,0), 1, cv2.LINE_AA)
        cv2.putText(image, str(counter),
                    (30,60),
                    cv2.FONT_HERSHEY_SIMPLEX, 1.3, (255,255,255), 2, cv2.LINE_AA)


        # Posture data
        cv2.putText(image, 'POSTURE', (70,90),
                    cv2.FONT_HERSHEY_SIMPLEX, 0.5, (0,0,0), 1, cv2.LINE_AA)
        cv2.putText(image, grip, (15,115),
                    cv2.FONT_HERSHEY_SIMPLEX, 0.6, (255,255,255), 1, cv2.LINE_AA)
        cv2.putText(image, stance, (15,140),
                    cv2.FONT_HERSHEY_SIMPLEX, 0.6, (255,255,255), 1, cv2.LINE_AA)

        # Stage data

        cv2.putText(image, 'STAGE', (145,15), cv2.FONT_HERSHEY_SIMPLEX, 0.5, (0, 0, 0), 1, cv2.LINE_AA)
        cv2.putText(image, stage, (130,45), cv2.FONT_HERSHEY_SIMPLEX, 1, (255, 255, 255), 2, cv2.LINE_AA)

        # Display Probability
        cv2.putText(image, f'CONF:{str(round(body_language_prob[np.argmax(body_language_prob)],2))}', (130,68), cv2.FONT_HERSHEY_SIMPLEX, 0.5, (0, 0, 0), 1, cv2.LINE_AA)

        # Render detections
        mp_drawing.draw_landmarks(image, results.pose_landmarks, mp_pose.POSE_CONNECTIONS,
                                mp_drawing.DrawingSpec(color=(0,30,0), thickness=2, circle_radius=2),
                                mp_drawing.DrawingSpec(color=(187,225,160), thickness=2, circle_radius=2)
                                 )

        cv2.imshow('NoRep app', image)

        if cv2.waitKey(10) & 0xFF == ord('q'):
            break

    cap.release()
    cv2.destroyAllWindows()