Research on the Construction of Human-Computer Interaction System Based on a Machine Learning Algorithm

Algorithm 1

Significance feature extraction algorithm based on bidirectional reference frame search.

Input: depth map sequence

Step 1. the initial feature point extraction

Use SURF algorithm to detect key points.

Step 2. Forward search

Initialize the reference frame of frame 1,

For do.

If then

end

Step 3. Backward search

Consistent with the forward search step, get the backward reference frame

Two-way search area fusion to obtain the final saliency area

Step 4. Feature selection and descriptor extraction

Use S to filter out invalid points, and extract HOG and HOF feature descriptions in a square area centered on key points .

Output: extracted feature points and location information and feature description