Research And Analysis On Human Visual System Interest Objects Detection

When observing a complex scene, the region of interest can obtain the most attention of the human visual system. Life is full of the redundant information and the human visual system can accurately capture the most interesting thing and ignore the others. People can establish their relation based on their common interests and hobbies. The interest object detection has been used in many aspects.With the use of computer simulation, the interest object detection models have been improved on precision and accuracy. However, there still has some limitations in the state-of-the-art model, such as the false positive problem in recognition, the low contrast in the region of interest, the lack of modeling-standard and so on. In order to solve the false positive problem in the existing interest object detection models, an improved algorithm based on multi-color information fusion (MCIF) has been proposed to overcome the above challenge. First of all, extracting the feature information in LMS space and the orientation information in RGB space. Then, the multi-color information fusion algorithm used to fuse the feature information. Finally, extracting the local significances to increase the details, and the interest regions are highlighted.Experiments have demonstrated the practical value of the proposed algorithm for the interest objects detection.