A Novel Window Median Filtering Algorithm for Real-Time Images
Dai Wen-gang · 2009
The matching performance of real-time image is an important factor for scene matching precision. However, during the acquisition of real-time images, some factors such as noise disturbance reduce the matching performance. Based on analysis of classical median filter and neighborhood averaging filter characteristics, a window median filter is designed, and apphed to real-time image preprocessing of scene matching. Filtering experiments indicate that the algorithm can increase the SNR of real-time image and improve the matching performance.