Hi!
I understand that you would like to track an object across multiple images (frames from a video).
In order to track an object, you would need to first detect it. As you mentioned, you can use feature extraction algorithms like SIFT and SIRF. You can also try out object detection based neural networks like the YOLO category of networks.
Once you have the object detections, you can track them using various trackers available in MATLAB like JPDA, GNN, TOMHT, etc.
Here is a MATLAB example you can go through that tracks vehicles from a video feed:
For more information on YOLO, JPDA, GNN and TOMHT, refer the following documentation links:
- https://in.mathworks.com/help/vision/ug/getting-started-with-yolo-v3.html?searchHighlight=yolo&s_tid=srchtitle_support_results_4_yolo
- https://in.mathworks.com/help/fusion/ref/trackergnn-system-object.html
- https://in.mathworks.com/help/fusion/ref/trackerjpda-system-object.html
- https://in.mathworks.com/help/fusion/ref/trackertomht-system-object.html
Hope this helps!
