Yolovideo / app.py
Dorn4449's picture
Update app.py
244fd38 verified
import cv2
import ultralytics
import gradio as gr
import torch
# Load YOLOv8 model
model = ultralytics.YOLO('yolov8n.pt')
# Set the stream URL (optional, can be commented out if using only camera)
stream_url = "https://edge01.london.nginx.hdontap.com/hosb5/ng_showcase-coke_bottle-street_fixed.stream/chunklist_w464099566.m3u8"
# Low-resolution for inference
LOW_RES = (320, 180)
def detect_and_draw(frame):
# Resize frame to low resolution for faster inference
low_res_frame = cv2.resize(frame, LOW_RES)
# Perform YOLOv8 inference
results = model(low_res_frame)
# Scale bounding boxes
scale_x = frame.shape[1] / LOW_RES[0]
scale_y = frame.shape[0] / LOW_RES[1]
# Draw bounding boxes on high-res frame
for detection in results[0].boxes.data:
x1, y1, x2, y2, conf, cls = detection
x1, y1, x2, y2 = int(x1*scale_x), int(y1*scale_y), int(x2*scale_x), int(y2*scale_y)
label = f"{results[0].names[int(cls)]} {conf:.2f}"
cv2.rectangle(frame, (x1, y1), (x2, y2), (0, 255, 0), 2)
cv2.putText(frame, label, (x1, y1 - 10), cv2.FONT_HERSHEY_SIMPLEX, 0.9, (0, 255, 0), 2)
return frame
def process_stream(stream_source):
if stream_source == "camera":
cap = cv2.VideoCapture(0) # Use 0 for the default camera
else:
cap = cv2.VideoCapture(stream_url)
frame_count = 0
while cap.isOpened():
ret, frame = cap.read()
if not ret:
break
frame_count += 3
if frame_count % 30 == 0: # Process every 30th frame
result = detect_and_draw(frame)
result_rgb = cv2.cvtColor(result, cv2.COLOR_BGR2RGB)
yield result_rgb
cap.release()
# Gradio interface for live video stream
iface = gr.Interface(
fn=process_stream,
inputs=gr.Dropdown(["camera", "stream"], label="Video Source"),
outputs="image",
live=True,
title="YOLOv8 Real-Time Object Detection",
description="Live stream processed with YOLOv8 for real-time object detection.")
if __name__ == "__main__":
if torch.cuda.is_available():
model.to('cuda')
iface.queue()
iface.launch()