Research on image processing application of improved adaptive filter based on LPSO algorithm

The detection technology based on image processing is a kind of nondestructive detection technology, which has the advantages of fast, safe, and high accuracy. It is necessary to use a filter to extract features in the process of detection. In this context, this paper studies the adaptive Log-Gabor filter improved on Log-Gabor filter based on the LPSO algorithm. The main research contents are as follows: Aiming at the problem that traditional Log-Gabor filter needs to select parameters and obtain the optimal solution is relatively troublesome, an adaptive Log-Gabor filter is proposed. PSO algorithm is used to improve it, but PSO algorithm is easy to fall into the local optimal solution, so Levy flight strategy is used to improve PSO algorithm, the improved algorithm is called LPSO algorithm, this method can make PSO algorithm successfully jump out of the local optimal solution. The LPSO algorithm is combined with a Log-Gabor filter to generate an adaptive Log-Gabor filter. The two adaptive filters are compared by simulation experiments.