Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
@@ -4,6 +4,7 @@ import torch
|
|
4 |
from ultralytics import YOLO
|
5 |
import gradio as gr
|
6 |
from scipy.interpolate import interp1d
|
|
|
7 |
import uuid
|
8 |
import os
|
9 |
import tempfile
|
@@ -12,7 +13,7 @@ import tempfile
|
|
12 |
model = YOLO("best.pt")
|
13 |
model.to('cuda' if torch.cuda.is_available() else 'cpu')
|
14 |
|
15 |
-
#
|
16 |
ball_class_index = None
|
17 |
for k, v in model.names.items():
|
18 |
if v.lower() == "cricketball":
|
@@ -25,7 +26,7 @@ if ball_class_index is None:
|
|
25 |
STUMPS_WIDTH = 0.2286
|
26 |
BALL_DIAMETER = 0.073
|
27 |
FRAME_RATE = 20
|
28 |
-
SLOW_MOTION_FACTOR =
|
29 |
CONF_THRESHOLD = 0.2
|
30 |
IMPACT_ZONE_Y = 0.85
|
31 |
IMPACT_DELTA_Y = 50
|
@@ -33,26 +34,6 @@ PITCH_LENGTH = 20.12
|
|
33 |
STUMPS_HEIGHT = 0.71
|
34 |
MAX_POSITION_JUMP = 30
|
35 |
|
36 |
-
def bezier_curve(points, num=100):
|
37 |
-
"""Compute a quadratic Bezier curve from given points."""
|
38 |
-
points = np.array(points)
|
39 |
-
if len(points) < 3:
|
40 |
-
return points
|
41 |
-
p0, p1, p2 = points[0], points[len(points)//2], points[-1]
|
42 |
-
t = np.linspace(0, 1, num=num)
|
43 |
-
curve = (1 - t)[:, None]**2 * p0 + 2 * (1 - t)[:, None] * t[:, None] * p1 + t[:, None]**2 * p2
|
44 |
-
return curve
|
45 |
-
|
46 |
-
def draw_stumps_overlay(frame):
|
47 |
-
height, width = frame.shape[:2]
|
48 |
-
stumps_x = width // 2
|
49 |
-
stump_top = int(height * 0.1)
|
50 |
-
stump_bottom = int(height * 0.9)
|
51 |
-
stump_width_px = int(width * 0.03)
|
52 |
-
for offset in [-stump_width_px, 0, stump_width_px]:
|
53 |
-
cv2.line(frame, (stumps_x + offset, stump_top), (stumps_x + offset, stump_bottom), (0, 0, 0), 2)
|
54 |
-
return frame
|
55 |
-
|
56 |
def process_video(video_path):
|
57 |
if not os.path.exists(video_path):
|
58 |
return [], [], [], "Error: Video file not found"
|
@@ -61,6 +42,7 @@ def process_video(video_path):
|
|
61 |
frame_height = int(cap.get(cv2.CAP_PROP_FRAME_HEIGHT))
|
62 |
frames, ball_positions, detection_frames, debug_log = [], [], [], []
|
63 |
frame_count = 0
|
|
|
64 |
while cap.isOpened():
|
65 |
ret, frame = cap.read()
|
66 |
if not ret:
|
@@ -80,11 +62,23 @@ def process_video(video_path):
|
|
80 |
frames[-1] = frame
|
81 |
debug_log.append(f"Frame {frame_count}: {detections} ball detections")
|
82 |
cap.release()
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
83 |
return frames, ball_positions, detection_frames, "\n".join(debug_log)
|
84 |
|
85 |
def find_bounce_point(ball_coords):
|
|
|
|
|
|
|
|
|
86 |
y_coords = [p[1] for p in ball_coords]
|
87 |
min_index = None
|
|
|
88 |
for i in range(2, len(y_coords) - 2):
|
89 |
dy1 = y_coords[i] - y_coords[i - 1]
|
90 |
dy2 = y_coords[i + 1] - y_coords[i]
|
@@ -92,70 +86,89 @@ def find_bounce_point(ball_coords):
|
|
92 |
if i > len(y_coords) * 0.2:
|
93 |
min_index = i
|
94 |
break
|
95 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
96 |
|
97 |
def estimate_trajectory(ball_positions, detection_frames, frame_height, frame_width):
|
98 |
if len(ball_positions) < 2:
|
99 |
return None, None, None, "Error: Not enough ball detections"
|
|
|
100 |
filtered_positions = [ball_positions[0]]
|
101 |
filtered_frames = [detection_frames[0]]
|
102 |
for i in range(1, len(ball_positions)):
|
103 |
-
|
104 |
-
|
|
|
105 |
filtered_frames.append(detection_frames[i])
|
|
|
106 |
if len(filtered_positions) < 2:
|
107 |
return None, None, None, "Error: Filtered detections too few"
|
|
|
108 |
x_vals = [p[0] for p in filtered_positions]
|
109 |
y_vals = [p[1] for p in filtered_positions]
|
110 |
times = np.array(filtered_frames) / FRAME_RATE
|
|
|
111 |
try:
|
112 |
fx = interp1d(times, x_vals, kind='cubic', fill_value="extrapolate")
|
113 |
fy = interp1d(times, y_vals, kind='cubic', fill_value="extrapolate")
|
114 |
except Exception as e:
|
115 |
return None, None, None, f"Interpolation error: {str(e)}"
|
116 |
-
|
117 |
-
|
|
|
|
|
|
|
118 |
trajectory = list(zip(x_full, y_full))
|
|
|
119 |
pitch_point = find_bounce_point(filtered_positions)
|
120 |
impact_point = filtered_positions[-1]
|
121 |
-
return trajectory, pitch_point, impact_point, "Trajectory estimated successfully"
|
122 |
|
123 |
-
|
124 |
-
if not frames or not trajectory or len(ball_positions) < 2:
|
125 |
-
return "Not enough data", trajectory, pitch_point, impact_point
|
126 |
-
frame_height, frame_width = frames[0].shape[:2]
|
127 |
-
stumps_x = frame_width / 2
|
128 |
-
stumps_y = frame_height * 0.9
|
129 |
-
stumps_width_pixels = frame_width * (STUMPS_WIDTH / 3.0)
|
130 |
-
pitch_x, _ = pitch_point
|
131 |
-
impact_x, _ = impact_point
|
132 |
-
if pitch_x < stumps_x - stumps_width_pixels / 2 or pitch_x > stumps_x + stumps_width_pixels / 2:
|
133 |
-
return f"Not Out (Pitched outside line)", trajectory, pitch_point, impact_point
|
134 |
-
if impact_x < stumps_x - stumps_width_pixels / 2 or impact_x > stumps_x + stumps_width_pixels / 2:
|
135 |
-
return f"Not Out (Impact outside line)", trajectory, pitch_point, impact_point
|
136 |
-
for x, y in trajectory:
|
137 |
-
if abs(x - stumps_x) < stumps_width_pixels / 2 and abs(y - stumps_y) < frame_height * 0.1:
|
138 |
-
return f"Out (Ball projected to hit stumps)", trajectory, pitch_point, impact_point
|
139 |
-
return f"Not Out (Missing stumps)", trajectory, pitch_point, impact_point
|
140 |
|
141 |
def generate_replay(frames, trajectory, pitch_point, impact_point, detection_frames):
|
142 |
if not frames or not trajectory:
|
143 |
return None
|
144 |
-
|
145 |
-
|
146 |
height, width = frames[0].shape[:2]
|
147 |
-
out = cv2.VideoWriter(
|
148 |
-
|
|
|
|
|
149 |
total_frames = max_frame - min_frame + 1
|
150 |
-
traj_per_frame = max(1, len(
|
151 |
-
indices = [min(i * traj_per_frame, len(
|
|
|
152 |
for i, frame in enumerate(frames):
|
153 |
-
frame = draw_stumps_overlay(frame)
|
154 |
idx = i - min_frame
|
155 |
if 0 <= idx < len(indices):
|
156 |
end_idx = indices[idx]
|
157 |
-
|
158 |
-
cv2.polylines(frame, [
|
159 |
if pitch_point and i == detection_frames[0]:
|
160 |
cv2.circle(frame, tuple(map(int, pitch_point)), 6, (0, 0, 255), -1)
|
161 |
if impact_point and i == detection_frames[-1]:
|
@@ -163,18 +176,21 @@ def generate_replay(frames, trajectory, pitch_point, impact_point, detection_fra
|
|
163 |
for _ in range(SLOW_MOTION_FACTOR):
|
164 |
out.write(frame)
|
165 |
out.release()
|
166 |
-
return
|
167 |
|
168 |
def drs_review(video):
|
169 |
frames, ball_positions, detection_frames, debug_log = process_video(video)
|
170 |
if not frames or not ball_positions:
|
171 |
return "No frames or detections found.", None
|
|
|
172 |
frame_height, frame_width = frames[0].shape[:2]
|
173 |
trajectory, pitch_point, impact_point, log = estimate_trajectory(ball_positions, detection_frames, frame_height, frame_width)
|
174 |
if not trajectory:
|
175 |
return f"{log}\n{debug_log}", None
|
|
|
176 |
decision, _, _, _ = lbw_decision(ball_positions, trajectory, frames, pitch_point, impact_point)
|
177 |
replay_path = generate_replay(frames, trajectory, pitch_point, impact_point, detection_frames)
|
|
|
178 |
result_log = f"DRS Decision: {decision}\n\n{log}\n\n{debug_log}"
|
179 |
return result_log, replay_path
|
180 |
|
@@ -184,11 +200,11 @@ iface = gr.Interface(
|
|
184 |
inputs=gr.Video(label="Upload Cricket Delivery Video"),
|
185 |
outputs=[
|
186 |
gr.Textbox(label="DRS Result and Debug Info"),
|
187 |
-
gr.Video(label="Replay with
|
188 |
],
|
189 |
-
title="GullyDRS -
|
190 |
-
description="Upload a cricket delivery video.
|
191 |
)
|
192 |
|
193 |
if __name__ == "__main__":
|
194 |
-
iface.launch()
|
|
|
4 |
from ultralytics import YOLO
|
5 |
import gradio as gr
|
6 |
from scipy.interpolate import interp1d
|
7 |
+
import plotly.graph_objects as go
|
8 |
import uuid
|
9 |
import os
|
10 |
import tempfile
|
|
|
13 |
model = YOLO("best.pt")
|
14 |
model.to('cuda' if torch.cuda.is_available() else 'cpu')
|
15 |
|
16 |
+
# Dynamically resolve ball class index
|
17 |
ball_class_index = None
|
18 |
for k, v in model.names.items():
|
19 |
if v.lower() == "cricketball":
|
|
|
26 |
STUMPS_WIDTH = 0.2286
|
27 |
BALL_DIAMETER = 0.073
|
28 |
FRAME_RATE = 20
|
29 |
+
SLOW_MOTION_FACTOR = 3
|
30 |
CONF_THRESHOLD = 0.2
|
31 |
IMPACT_ZONE_Y = 0.85
|
32 |
IMPACT_DELTA_Y = 50
|
|
|
34 |
STUMPS_HEIGHT = 0.71
|
35 |
MAX_POSITION_JUMP = 30
|
36 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
37 |
def process_video(video_path):
|
38 |
if not os.path.exists(video_path):
|
39 |
return [], [], [], "Error: Video file not found"
|
|
|
42 |
frame_height = int(cap.get(cv2.CAP_PROP_FRAME_HEIGHT))
|
43 |
frames, ball_positions, detection_frames, debug_log = [], [], [], []
|
44 |
frame_count = 0
|
45 |
+
|
46 |
while cap.isOpened():
|
47 |
ret, frame = cap.read()
|
48 |
if not ret:
|
|
|
62 |
frames[-1] = frame
|
63 |
debug_log.append(f"Frame {frame_count}: {detections} ball detections")
|
64 |
cap.release()
|
65 |
+
|
66 |
+
if not ball_positions:
|
67 |
+
debug_log.append("No balls detected in any frame")
|
68 |
+
else:
|
69 |
+
debug_log.append(f"Total ball detections: {len(ball_positions)}")
|
70 |
+
debug_log.append(f"Video resolution: {frame_width}x{frame_height}")
|
71 |
+
|
72 |
return frames, ball_positions, detection_frames, "\n".join(debug_log)
|
73 |
|
74 |
def find_bounce_point(ball_coords):
|
75 |
+
"""
|
76 |
+
Detect bounce point using y-derivative reversal with early-frame suppression.
|
77 |
+
Looks for where y increases then decreases (ball hits ground).
|
78 |
+
"""
|
79 |
y_coords = [p[1] for p in ball_coords]
|
80 |
min_index = None
|
81 |
+
|
82 |
for i in range(2, len(y_coords) - 2):
|
83 |
dy1 = y_coords[i] - y_coords[i - 1]
|
84 |
dy2 = y_coords[i + 1] - y_coords[i]
|
|
|
86 |
if i > len(y_coords) * 0.2:
|
87 |
min_index = i
|
88 |
break
|
89 |
+
|
90 |
+
if min_index is not None:
|
91 |
+
return ball_coords[min_index]
|
92 |
+
|
93 |
+
return ball_coords[len(ball_coords)//2]
|
94 |
+
|
95 |
+
def lbw_decision(ball_positions, trajectory, frames, pitch_point, impact_point):
|
96 |
+
if not frames or not trajectory or len(ball_positions) < 2:
|
97 |
+
return "Not enough data", trajectory, pitch_point, impact_point
|
98 |
+
|
99 |
+
frame_height, frame_width = frames[0].shape[:2]
|
100 |
+
stumps_x = frame_width / 2
|
101 |
+
stumps_y = frame_height * 0.9
|
102 |
+
stumps_width_pixels = frame_width * (STUMPS_WIDTH / 3.0)
|
103 |
+
|
104 |
+
pitch_x, _ = pitch_point
|
105 |
+
impact_x, impact_y = impact_point
|
106 |
+
|
107 |
+
if pitch_x < stumps_x - stumps_width_pixels / 2 or pitch_x > stumps_x + stumps_width_pixels / 2:
|
108 |
+
return f"Not Out (Pitched outside line)", trajectory, pitch_point, impact_point
|
109 |
+
if impact_x < stumps_x - stumps_width_pixels / 2 or impact_x > stumps_x + stumps_width_pixels / 2:
|
110 |
+
return f"Not Out (Impact outside line)", trajectory, pitch_point, impact_point
|
111 |
+
for x, y in trajectory:
|
112 |
+
if abs(x - stumps_x) < stumps_width_pixels / 2 and abs(y - stumps_y) < frame_height * 0.1:
|
113 |
+
return f"Out (Ball projected to hit stumps)", trajectory, pitch_point, impact_point
|
114 |
+
return f"Not Out (Missing stumps)", trajectory, pitch_point, impact_point
|
115 |
|
116 |
def estimate_trajectory(ball_positions, detection_frames, frame_height, frame_width):
|
117 |
if len(ball_positions) < 2:
|
118 |
return None, None, None, "Error: Not enough ball detections"
|
119 |
+
|
120 |
filtered_positions = [ball_positions[0]]
|
121 |
filtered_frames = [detection_frames[0]]
|
122 |
for i in range(1, len(ball_positions)):
|
123 |
+
prev, curr = filtered_positions[-1], ball_positions[i]
|
124 |
+
if np.linalg.norm(np.array(curr) - np.array(prev)) <= MAX_POSITION_JUMP:
|
125 |
+
filtered_positions.append(curr)
|
126 |
filtered_frames.append(detection_frames[i])
|
127 |
+
|
128 |
if len(filtered_positions) < 2:
|
129 |
return None, None, None, "Error: Filtered detections too few"
|
130 |
+
|
131 |
x_vals = [p[0] for p in filtered_positions]
|
132 |
y_vals = [p[1] for p in filtered_positions]
|
133 |
times = np.array(filtered_frames) / FRAME_RATE
|
134 |
+
|
135 |
try:
|
136 |
fx = interp1d(times, x_vals, kind='cubic', fill_value="extrapolate")
|
137 |
fy = interp1d(times, y_vals, kind='cubic', fill_value="extrapolate")
|
138 |
except Exception as e:
|
139 |
return None, None, None, f"Interpolation error: {str(e)}"
|
140 |
+
|
141 |
+
total_frames = max(filtered_frames) - min(filtered_frames) + 1
|
142 |
+
t_full = np.linspace(times[0], times[-1], max(5, total_frames * SLOW_MOTION_FACTOR))
|
143 |
+
x_full = fx(t_full)
|
144 |
+
y_full = fy(t_full)
|
145 |
trajectory = list(zip(x_full, y_full))
|
146 |
+
|
147 |
pitch_point = find_bounce_point(filtered_positions)
|
148 |
impact_point = filtered_positions[-1]
|
|
|
149 |
|
150 |
+
return trajectory, pitch_point, impact_point, "Trajectory estimated successfully"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
151 |
|
152 |
def generate_replay(frames, trajectory, pitch_point, impact_point, detection_frames):
|
153 |
if not frames or not trajectory:
|
154 |
return None
|
155 |
+
|
156 |
+
temp_file = os.path.join(tempfile.gettempdir(), f"drs_output_{uuid.uuid4()}.mp4")
|
157 |
height, width = frames[0].shape[:2]
|
158 |
+
out = cv2.VideoWriter(temp_file, cv2.VideoWriter_fourcc(*'mp4v'), FRAME_RATE / SLOW_MOTION_FACTOR, (width, height))
|
159 |
+
|
160 |
+
min_frame = min(detection_frames)
|
161 |
+
max_frame = max(detection_frames)
|
162 |
total_frames = max_frame - min_frame + 1
|
163 |
+
traj_per_frame = max(1, len(trajectory) // total_frames)
|
164 |
+
indices = [min(i * traj_per_frame, len(trajectory)-1) for i in range(total_frames)]
|
165 |
+
|
166 |
for i, frame in enumerate(frames):
|
|
|
167 |
idx = i - min_frame
|
168 |
if 0 <= idx < len(indices):
|
169 |
end_idx = indices[idx]
|
170 |
+
points = np.array(trajectory[:end_idx+1], dtype=np.int32).reshape((-1, 1, 2))
|
171 |
+
cv2.polylines(frame, [points], False, (255, 0, 0), 2)
|
172 |
if pitch_point and i == detection_frames[0]:
|
173 |
cv2.circle(frame, tuple(map(int, pitch_point)), 6, (0, 0, 255), -1)
|
174 |
if impact_point and i == detection_frames[-1]:
|
|
|
176 |
for _ in range(SLOW_MOTION_FACTOR):
|
177 |
out.write(frame)
|
178 |
out.release()
|
179 |
+
return temp_file
|
180 |
|
181 |
def drs_review(video):
|
182 |
frames, ball_positions, detection_frames, debug_log = process_video(video)
|
183 |
if not frames or not ball_positions:
|
184 |
return "No frames or detections found.", None
|
185 |
+
|
186 |
frame_height, frame_width = frames[0].shape[:2]
|
187 |
trajectory, pitch_point, impact_point, log = estimate_trajectory(ball_positions, detection_frames, frame_height, frame_width)
|
188 |
if not trajectory:
|
189 |
return f"{log}\n{debug_log}", None
|
190 |
+
|
191 |
decision, _, _, _ = lbw_decision(ball_positions, trajectory, frames, pitch_point, impact_point)
|
192 |
replay_path = generate_replay(frames, trajectory, pitch_point, impact_point, detection_frames)
|
193 |
+
|
194 |
result_log = f"DRS Decision: {decision}\n\n{log}\n\n{debug_log}"
|
195 |
return result_log, replay_path
|
196 |
|
|
|
200 |
inputs=gr.Video(label="Upload Cricket Delivery Video"),
|
201 |
outputs=[
|
202 |
gr.Textbox(label="DRS Result and Debug Info"),
|
203 |
+
gr.Video(label="Replay with Trajectory & Decision")
|
204 |
],
|
205 |
+
title="GullyDRS - AI-Powered LBW Review",
|
206 |
+
description="Upload a cricket delivery video. The system will track the ball, estimate trajectory, and return a replay with an OUT/NOT OUT decision."
|
207 |
)
|
208 |
|
209 |
if __name__ == "__main__":
|
210 |
+
iface.launch()
|