The image positioning of the thick fog people trapped in Based on the improved median filter
Honghua Wang, Hui Zhang · 2014 IEEE Workshop on Advanced Research and Technology in Industry Applications (WARTIA) · 2014
Traditional methods for positioning accuracy is not high and takes long time consuming problem. A real-time positioning algorithm is proposed based on improved median filter people trapped in the thick fog. first of all, Visual zoning of heavy fog environment is initialized, after initialization operation, to average filtering processing of heavy fog environment, and the heavy fog in the environment is not peer pixel jump frequency statistics, jump times lower than 15 of the heavy fog image is the background region, over 15 line is stranded the personnel of the pixel area. And introducing spatial location parameters, calculate the location coordinates of people trapped in the thick fog, the prior edge detection method with the aid of edge number of regional characteristics gets the accurate location of people trapped in thick fog. The experimental results shows that the improved method than the traditional positioning method is high positioning accuracy, short time consuming, strong real-time performance and broad application space.