Image Contour Detection Inspired By Biological Vision

Images are a way of characterizing the objective world.Because images are simple and easy to understand,they are widely chosen as information eariers.The biological vision system is extremely efficient in processing visual information such as images.Through in-depth study of its internal mechanism,it will greatly help humans to establish image processing methods based on biological vision mechanisms.In this paper,the biological vision mechanism is fully considered.Based on the hierarchical response mechanism of the biological vision mechanism,a fast contour detection method based on the primary visual pathway hierarchical response model is constructed.Considering the visual information transmission and processing flow in the primary visual pathway.A contour detection method based on the primary vision path computation model is proposed.Considering the processing mechanism of visual information in the biological vision system,a contour detection method based on the interaction of visual information level is proposed.The most important achievements and work of this paper are as follows:(1)Proposed a new model of hierarchical transmission and response of primary visual path information.In the cortical visual pathway,the CRF and NCRF feelings are reflected in the processing and adjustment mechanism of visual information,and the hierarchical correlation characteristics of different types of receptive fields in extracting visual information are considered.Cortical lower visual pathways are also introduced for visual information.The holistic description proposes a fusion strategy for multi-view visual response,which achieves the synergy between local detection and holistic perception.The RuG40 library were taken for processing,the optimal average indicators for the entire data set and single map of the RuG40 library are 0.46 and 0.48,respectively.(2)Proposed a new contour detection method based on the primary vision path computation model.This paper attempts to construct a primary vision path computation model to simulate the transmission and processing of visual information flows.First,in the retinal ganglion,a classical receptive field direction selection model that combines multi-scale features is constructed to obtain the primary contour response of the image target.Then,a spatiotemporal coding mechanism is utilized to streamline the redundant features in the primary contour response,in the visual pathway of the retinal ganglion to the LGN.Furthermore,the synergistic effect of NSCT and Gabor transform is used to simulate the processing effect of the lateral suppression characteristics of NCRF on the texture background information.Finally,the feedforward mechanism of the visual pathway to the primary visual cortex is combined with the visual features of the multi-visual pathway to obtain the contour response.The RuG40 library were taken for processing,the optimal average indicators for the entire data set and single map of the RuG40 library are 0.49 and 0.56,respectively.Compared with the ISO and MCI methods,this method has a significant improvement in the accuracy of contour integrity and texture background suppression.(3)Proposed a new contour detection method based on interaction mechanism between visual information flows.Based on the primary visual pathway computation model constructed in Chapter 4,the frequency domain separation model based on NSCT is constructed to realize the frequency domain separation of visual information by LGN.Constructed a plurality of rapid visual information transfer path to simulate the interaction between the visual information hierarchy.A multi-level visual feature fusion model of the primary visual cortex was constructed to achieve rapid fusion of multi-level visual information.Considering the complex texture background of medical images,this chapter selects two typical tasks of medical image processing-colony detection and white blood cell detection as application projects to verify the effectiveness of the algorithm constructed in this chapter.The experimental results show that the proposed method can not only preserve the contour of the target body more completely,but also suppress the texture background better,which lays a good foundation for the subsequent operation.