基于视觉注意机制的图像分割研究及其应用-计算机应用技术专业论文.docxVIP

RESEARCH AND APPLICATION OF IMAGE SEGMENTATION BASED ON VISUAL ATTENTION MECHANISM ABSTRACT Image segmentation is an key technology of image processing and computer vision. It helps to improve the veracity of certain target’s location based on image content, and techniques like image editing cutout can not be separated from the correct segmentation. Image segmentation method voluminous, but to achieve a universal segmentation technology is also faced with great difficulties. The method of image segmentation is tremendous, but to achieve a universal segmentation technology is also faced with great difficulties.With the integrated development of computer vision, modern physiology, neuro-psychology, object recognition, image processing and so many other disciplines, image segmentation technique based on visual attention is becoming increasingly attracted peoples attention. It belongs to the international advanced research topics, Its theoretical results have great contribution to intelligent research and develo- pment. It has broad application prospects in Meteorological services in remote sensing, medical image analysis, machine manufacturing, product testing, military research, transportation image analysis and other areas. The main content of the paper is to establish a visual attention computational model that has good capacity of quantitative description, using it to guide the image segmentation, to actualize objects’ tracking or location. The paper describes the biological basis of visual attention mechani- sm first, the key technologies of the Itti visual attention computational model are introduced in detail, and conducted experiments, found that there are many deficiencies in Itti visual attention computational model, it has pseudo attention point, miss detection, the focus of attention’s track- ing process is unreasonable and lack integrity description ability of target. The paper has been improved the Itti visual attention computational model, mainly on h