File size: 5,445 Bytes
2649d00
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
import cv2
import numpy as np
import gradio as gr
import tempfile
from pathlib import Path
from cvzone.ColorModule import ColorFinder
from batsman import batsman_detect
from ball_detect import ball_detect

# ---------------------------------------------------
# Static detection parameters (tune these if needed)
# ---------------------------------------------------
mycolorFinder = ColorFinder(False)

# HSV range for ball – tune for your footage
hsvVals = {
    "hmin": 10,
    "smin": 44,
    "vmin": 192,
    "hmax": 125,
    "smax": 114,
    "vmax": 255,
}

# RGB range & Canny thresholds for batsman – replace with tuned values
tuned_rgb_lower = np.array([112, 0, 181])
tuned_rgb_upper = np.array([255, 255, 255])
tuned_canny_threshold1 = 100
tuned_canny_threshold2 = 200

# ---------------------------------------------------
# Helper to classify each frame event
# ---------------------------------------------------

def ball_pitch_pad(x, x_prev, prev_x_diff, y, y_prev, prev_y_diff, batLeg):
    """Return 'Pad', 'Pitch' or 'Motion' based on ball & batsman coords."""
    if x_prev == 0 and y_prev == 0:
        return "Motion", 0, 0

    if abs(x - x_prev) > 3 * abs(prev_x_diff) and abs(prev_x_diff) > 0:
        if y < batLeg:
            return "Pad", x - x_prev, y - y_prev

    if y - y_prev < 0 and prev_y_diff > 0:
        if y < batLeg:
            return "Pad", x - x_prev, y - y_prev
        else:
            return "Pitch", x - x_prev, y - y_prev

    return "Motion", x - x_prev, y - y_prev

# ---------------------------------------------------
# Main analysis routine wrapped for Gradio
# ---------------------------------------------------

def detect_lbw(video):
    """Run the LBW detector on an uploaded video, return annotated clip + verdict."""
    # Accept both dict (gradio 4.x) or str
    video_path = video if isinstance(video, (str, Path)) else video.get("name")

    cap = cv2.VideoCapture(str(video_path))
    if not cap.isOpened():
        raise ValueError("Unable to open uploaded video.")

    fps = cap.get(cv2.CAP_PROP_FPS) or 25
    width = int(cap.get(cv2.CAP_PROP_FRAME_WIDTH))
    height = int(cap.get(cv2.CAP_PROP_FRAME_HEIGHT))

    # Prepare temporary output file for annotated video
    tmpfile = tempfile.NamedTemporaryFile(delete=False, suffix=".mp4")
    out_path = tmpfile.name
    fourcc = cv2.VideoWriter_fourcc(*"mp4v")
    writer = cv2.VideoWriter(out_path, fourcc, fps, (width, height))

    # Tracking variables
    x = y = batLeg = 0
    x_prev = y_prev = 0
    prev_x_diff = prev_y_diff = 0
    lbw_detected = False

    while True:
        x_prev, y_prev = x, y
        success, img = cap.read()
        if not success:
            break

        overlay = img.copy()

        # 1️⃣ Ball detection
        _, x, y = ball_detect(img, mycolorFinder, hsvVals)
        if x and y:
            cv2.circle(overlay, (x, y), 8, (255, 0, 0), -1)

        # 2️⃣ Batsman detection
        batsmanContours = batsman_detect(
            img,
            tuned_rgb_lower,
            tuned_rgb_upper,
            tuned_canny_threshold1,
            tuned_canny_threshold2,
        )

        # Compute batsman's leg (lowest y among contours above the ball)
        current_batLeg = float("inf")
        for cnt in batsmanContours:
            if cv2.contourArea(cnt) > 5000 and y != 0 and min(cnt[:, :, 1]) < y:
                leg_candidate = max(cnt[:, :, 1])
                current_batLeg = min(current_batLeg, leg_candidate)
                cv2.drawContours(overlay, cnt, -1, (0, 255, 0), 3)
        batLeg = current_batLeg if current_batLeg != float("inf") else batLeg

        # 3️⃣ Classify the motion event for this frame
        motion_type, prev_x_diff, prev_y_diff = ball_pitch_pad(
            x, x_prev, prev_x_diff, y, y_prev, prev_y_diff, batLeg
        )

        if motion_type == "Pad":
            lbw_detected = True
            cv2.putText(
                overlay,
                "PAD CONTACT",
                (50, 80),
                cv2.FONT_HERSHEY_SIMPLEX,
                1.6,
                (0, 0, 255),
                4,
                cv2.LINE_AA,
            )
        elif motion_type == "Pitch":
            cv2.putText(
                overlay,
                "Bounced",
                (50, 80),
                cv2.FONT_HERSHEY_SIMPLEX,
                1.6,
                (0, 255, 255),
                4,
                cv2.LINE_AA,
            )

        writer.write(overlay)

    cap.release()
    writer.release()

    verdict = "✅ Potential LBW Detected!" if lbw_detected else "❌ No LBW Detected."
    return out_path, verdict

# ---------------------------------------------------
# Gradio interface
# ---------------------------------------------------

demo = gr.Interface(
    fn=detect_lbw,
    inputs=gr.Video(label="Upload cricket clip (side-on view)"),
    outputs=[
        gr.Video(label="Annotated Review"),
        gr.Textbox(label="Decision"),
    ],
    title="Automated LBW Detector",
    description=(
        "Upload a short video of the delivery. The system analyses the ball & batsman "
        "interaction frame-by-frame, overlays detections, and flags potential LBW instances."
    ),
)

if __name__ == "__main__":
    demo.launch()