Research on Key Technology for Processing Images from Infrared Trigger Cameras

摘要

In recent years, infrared trigger cameras are widely employed in the fields of wildlife monitoring and protection as they can collect massive image data. It is of great realistic significance about how to conduct automatic processing of these image data and extract key information. This paper has a detailed analysis about features of images in this type by automatically grouping them based on time intervals in order to eliminate the impact of light intensity on images in different time frames, automatically eliminating the background of those images within the groups and extracting the image area of individual animals (namely Region of Interest, short for ROI) and identifying the image features of these ROI areas in terms of species. The above-mentioned key technique has been tested for many times and achieved quite good results, but still has the potential of further optimization and improvement.