How object is detected by openvx kernel ?
How object is detected by openvx kernel ?[/QUOTE]
The answer depends on what type of object you’re trying to detect and the particulars of your application. If you can provide some details on what you’re planning to do, I may be able to give a better answer.
In the meantime, here are three possible avenues to consider for object detection. The first is based on the core OpenVX kernels, the other two approaches require OpenVX extensions that may or may not be supported in your OpenVX implementation.
1.) OpenVX offers a whole suite of traditional image processing filters that are useful for feature recognition (e.g. Harris/fast corners, Sobel, optical flow, etc.). Check the computer vision literature for guidance on how to arrange these filters in a manner suitable to your workload.
2.) The Classifier Extension provides a data/node type that supports traditional computer-vision classification models (SVMs, etc.)
3.) The Neural Network extension provides the building blocks to implement neural networks as OpenVX graphs.
I want to implement blind spot detection algorithm from open vx .