High-Speed-Tracking

Introduction

This repository provides a method to track a moving object and access its location in real time using the OpenCV template matching method. To ensure efficiency, the code runs in three threads:

1. The first thread collects frames from the camera.
2. The second thread detects the location of the object in each frame.
3. The third thread saves the object locations to a .npy file on your local device, allowing real-time access to the object locations.

How to Use the Code

Saving the Data

An example of how to use this code can be found in the main branch of this repository in three_threads.py.

1. Connect the Camera:

   • Connect the camera and adjust the settings to get a good image of the object being tracked.

2. Initialize Queues:

   • Start two queues: one to hold the frames from the camera, and another to hold the location of the object.

3. Load the Template:

   • Import the template using the load_template function.

4. Set Parameters:

   • Determine the number of frames to be captured and processed.
   • Decide how often location data will be saved locally.
   • Specify the file path for saving locations. If the file does not exist, it will be automatically created.

5. Create Threads:

   • Create each thread using threading.Thread(target=, args=()).

6. Start Threads:

   • Start each thread using .start().

7. Stop Threads:

   • Stop the threads using .join().

8. Disconnect the Camera:

   • Disconnect the camera when done.

Reading the Data in Real Time

To read the data while locations are being collected, use the read_locations function. An example is provided in how_to_access_locations.py.

Possible Issues

1. Overwriting Data:

   • If the file chosen to save the data already exists, the new data will overwrite the old data.

2. Image Format:

   • The code assumes that all images are in grayscale.

3. Batch Size:

   • The batch size must be a multiple of the number of frames to be collected.

4. Python Version:

   • The Baumer camera used in this project requires Python version 3.10.2.

This repository provides a comprehensive solution for real-time object tracking using OpenCV, ensuring efficient and effective tracking through multi-threading.