Gaussian Mixture Models implementation in Real time for Object Detection

Question

I am working on real time video compression. I am modeling it with MATLAB. I need to later implement it on DSP processor with 5832MIPS, 729MHz.

Is it feasible to implement Gaussian Mixture models on DSP processor or are there better algorithms for object detection ?

Thanks in advance

Answer 1

Any kind of iterative procedure, in particular mixture model fitting, is quite time-consuming, and thus unlikely to be fast enough for real-time processing, unless you have only a handful of kernels to fit (we fit hundreds to a few thousands of spots, which takes several seconds per frame, despite our fast mixture-model implementation).

Why don't you use a non-iterative spot detector, such as H-Dome? Or do you need sub-pixel positions of your features?