write a python script that uses yolo model for object detection that takes in image from…
- User
write a python script that uses yolo model for object detection that takes in image from my device
Assistant
Here's a complete and well-documented Python script that uses the YOLOv5 model (from the Ultralytics library) to perform object detection on an image from your device.
✅ Requirements
Before running the script, make sure you have the following installed:
pip install torch torchvision torchaudio pip install ultralytics⚠️ If you have a compatible GPU, install the appropriate PyTorch version from PyTorch's official website.
🧠 Script Overview
This script:
- Loads a pre-trained YOLOv5 model (
yolov5s). - Accepts an image path as a command-line argument.
- Performs object detection on the image.
- Displays the image with detected objects highlighted using bounding boxes and labels.
📄 Python Script:
yolo_detection.pyimport sys import cv2 from ultralytics import YOLO def main(): # Check if the image path is provided if len(sys.argv) < 2: print("Usage: python yolo_detection.py <image_path>") print("Example: python yolo_detection.py images/test.jpg") return image_path = sys.argv[1] # Load a pre-trained YOLOv5 model # 'yolov5s' is a lightweight version; you can use 'yolov5m', 'yolov5l', etc. model = YOLO('yolov5s') # Perform object detection on the image results = model.predict(source=image_path, conf=0.5) # 0.5 is the confidence threshold # Display the result for result in results: # Draw bounding boxes and labels on the image annotated_frame = result.plot() # Show the image with detections cv2.imshow('Object Detection', annotated_frame) cv2.waitKey(0) # Wait until a key is pressed cv2.destroyAllWindows() if __name__ == "__main__": main()
🖼️ How to Use
-
Place an image file (e.g.,
test.jpg) in the same directory as the script or provide the full path. -
Run the script from the terminal:
python yolo_detection.py images/test.jpgReplace
images/test.jpgwith the actual path to your image.
📝 Optional Enhancements
-
Save the output image:
result.save(filename="output.jpg") -
Adjust confidence threshold:
results = model.predict(source=image_path, conf=0.7) -
Add custom class filtering:
results = model.predict(source=image_path, classes=[0, 1, 2]) # Detect only specific classes
🧪 Sample Output
When you run the script, a window will open showing your image with:
- Bounding boxes around detected objects.
- Labels (e.g., "person", "car", "dog").
- Confidence scores (if enabled).
🧩 Notes
- YOLOv5 automatically downloads the model weights (
yolov5s.pt) if not found. - The script uses OpenCV (
cv2) for displaying the image. If you're using a headless environment, consider saving the result instead of displaying it.
Let me know if you'd like to extend this to video, webcam, or custom datasets!
- Loads a pre-trained YOLOv5 model (
Other models, same or similar prompt
Responses were produced in different, uncontrolled contexts (system prompts, settings, prior turns). This is not a benchmark.