Closing of Camera Module in offline recognition

[Jayashankar-K]

Search before asking

• I have searched the YOLOv5 issues and discussions and found no similar questions.

Question

I am working on a project which involves a condition based detection.
Which means only if a condition is met then i invoke the YOLOv5 to perform a detection from my live camera feed.
The entire code runs in an infinite loop
If the detection conditions are satisfied then the again it should check for the condition and perform detection again.

Note : I am running this on a Anaconda environment

MY ISSUE :
In the first iteration the camera module is generated and detection is performed.
After which the second iteration is unable to invoke and generating errors due to not closing of the camera module from the first iteration.

The camera module window is only closing if i press "Ctrl+C" , but all the programs running are also terminated and the conda environment returns to a static state.

Is there any way to kill only the generated camera module without stopping all the processes so that the program can run in a loop without any errors.

TRIED METHODS

1. Termination of subprocess using commands in Conda env
   I have written a code snippet to find the PID of the subprocess and tried to terminate it but it does not work.

2.Using os.kill(os.getpid(), 2)
This stopped execution of all the processes

3.Tried to use cap.close() and cv2.CloseAllWIndows()
This also does not work

I have also tried to make changes in detect.py and dataloaders.py but it was of no use.

SO IF ANYONE CAN POINT ME TO WHAT CAN BE THE POSSIBLE MISTAKE I AM DOING IT WOULD BE VERY HELPFUL.THANK YOU

Additional

I am ready to provide more information or code snippets if needed.

[github-actions]

👋 Hello @Jayashankar-K, thank you for your interest in YOLOv5 🚀! Please visit our ⭐️ Tutorials to get started, where you can find quickstart guides for simple tasks like Custom Data Training all the way to advanced concepts like Hyperparameter Evolution.

If this is a 🐛 Bug Report, please provide a minimum reproducible example to help us debug it.

If this is a custom training ❓ Question, please provide as much information as possible, including dataset image examples and training logs, and verify you are following our Tips for Best Training Results.

Requirements

Python>=3.8.0 with all requirements.txt installed including PyTorch>=1.8. To get started:

git clone https://github.com/ultralytics/yolov5  # clone
cd yolov5
pip install -r requirements.txt  # install

Environments

YOLOv5 may be run in any of the following up-to-date verified environments (with all dependencies including CUDA/CUDNN, Python and PyTorch preinstalled):

• Notebooks with free GPU:
• Google Cloud Deep Learning VM. See GCP Quickstart Guide
• Amazon Deep Learning AMI. See AWS Quickstart Guide
• Docker Image. See Docker Quickstart Guide

Status

If this badge is green, all YOLOv5 GitHub Actions Continuous Integration (CI) tests are currently passing. CI tests verify correct operation of YOLOv5 training, validation, inference, export and benchmarks on macOS, Windows, and Ubuntu every 24 hours and on every commit.

Introducing YOLOv8 🚀

We're excited to announce the launch of our latest state-of-the-art (SOTA) object detection model for 2023 - YOLOv8 🚀!

Designed to be fast, accurate, and easy to use, YOLOv8 is an ideal choice for a wide range of object detection, image segmentation and image classification tasks. With YOLOv8, you'll be able to quickly and accurately detect objects in real-time, streamline your workflows, and achieve new levels of accuracy in your projects.

Check out our YOLOv8 Docs for details and get started with:

pip install ultralytics

[glenn-jocher]

@Jayashankar-K hello! It sounds like you're facing an issue with properly releasing the camera resource between iterations of your detection loop. To ensure the camera is released correctly, you should explicitly release the capture object at the end of each detection cycle.

Here's a general approach you can follow:

1. Initialize the camera capture object (cap) outside of your loop.
2. Inside the loop, after the detection is done and before checking the condition again, release the camera using cap.release().
3. Also, make sure to call cv2.destroyAllWindows() to close any OpenCV windows that might be open.
4. Reinitialize the capture object when the condition is met again.

Your loop might look something like this:

cap = cv2.VideoCapture(0)  # Initialize outside the loop
while True:
    if condition_met:
        # Perform detection using cap
        # ...
        cap.release()  # Release after detection
        cv2.destroyAllWindows()  # Close any windows
        # Reinitialize if needed
        cap = cv2.VideoCapture(0)
    # Other conditions or sleep to prevent tight looping

Make sure that the cap.release() and cv2.destroyAllWindows() calls are placed correctly to ensure they are executed after each detection cycle.

If you continue to face issues, please provide a minimal code snippet that reproduces the problem, and we'll take a closer look. Remember to handle any exceptions and edge cases where the camera might not be initialized or released properly.

For more detailed guidance, you can refer to our documentation at https://docs.ultralytics.com/yolov5/.

Keep up the great work, and I hope this helps you resolve the issue! 😊

[Jayashankar-K]

@glenn-jocher I have tried your solution but it does not work.
I am able continuously run my main program for the first iteration but in the second iteration it gives out an error , as the camera module is not closing after the first iteration.
it is showing this error -

here are my codes if you want to take a look

1- After my initial condition is met then i invoke the detection algorithm
Its code https://www.ideone.com/X9qEIP

Then this code invokes detect.py program in which the major part of creating a camera module and detection process takes place.
detect.py code --> https://www.ideone.com/asGX9R

If i am able to successfully close the window then my project will be completed.

Thank you for your help and support.
I can provide with more information if needed.

[glenn-jocher]

@Jayashankar-K, I'm sorry to hear that the issue persists. Unfortunately, I cannot view the screenshots or external code links you've provided due to the guidelines of this platform. However, I can still offer some general advice.

The error message you're seeing typically indicates that the camera resource is not being released properly or that there's an issue reinitializing the camera capture after it's been released.

Here are a few steps you can take to troubleshoot the issue:

1. Ensure Proper Release: Double-check that cap.release() and cv2.destroyAllWindows() are being called at the end of each detection cycle, even if an exception occurs. Use a try...finally block to guarantee that these lines are executed:

try:
    # Your detection code
    pass
finally:
    cap.release()
    cv2.destroyAllWindows()

2. Check Camera Reinitialization: Make sure that the camera is not being accessed by another process and that it's properly reinitialized before starting a new detection cycle.

3. Error Handling: Add error handling around the camera capture code to catch any exceptions that might provide more insight into what's going wrong.

4. Minimal Reproducible Example: Since I can't view your code, try to create a minimal reproducible example that isolates the camera handling logic. This will help you identify if the issue is with the camera handling or another part of your code.

5. Camera Hardware: Test with a different camera or on a different system to rule out hardware-specific issues.

6. Logging: Add logging throughout your code to track the program's flow and pinpoint where the failure occurs.

If you continue to experience issues, please provide a minimal reproducible example directly in the GitHub issue, using code blocks to share snippets, so that I can review it within the constraints of this platform.

Remember, the goal is to ensure that the camera is released after each use and properly reinitialized before the next use. Keep iterating on this process, and you should be able to find a solution. Good luck! 😊

[Jayashankar-K]

I found a solution.
In my solution i initiated the detection algorithm with a thread and killed the thread after my program got executed.

[Jayashankar-K]

Thank you for your support and help. 💯 @glenn-jocher

[glenn-jocher]

@Jayashankar-K, that's fantastic news! I'm glad to hear you found a solution that works for your project. Using threading to manage the lifecycle of the detection algorithm and ensuring resources are properly released is a clever approach. Great job on troubleshooting and implementing a fix! 🎉

If you have any more questions or need further assistance as you continue to develop your project, feel free to reach out. The YOLOv5 community and the Ultralytics team are always here to help. Keep up the great work, and best of luck with your project! 😊