Adaptive Neighborhood Morphology Theory And Its FPGA Implementation

Mathematical morphology has a rigorous mathematical theoretical foundation and a streamlined and efficient thinking.Its basic theories and methods have been widely applied to all aspects of image processing.Mathematical morphology image processing is the use of structural elements as a probe to collect image information.For the same image,different structural elements are used for morphological operations,and different processing results are obtained.Therefore,the selection of structural elements is extremely large.Affects the effect of morphological image processing.However,classical mathematical morphology uses fixed structural elements in image processing,and can not adapt well to multi-target images or images with large target changes,which leads to a large number of image details,so adaptively select structures according to the characteristics of the images themselves.Elements are especially important.Firstly,aiming at the local features of the image and the position information of the pixels,this paper proposes a morphological operator that selects the structural elements according to the features of the neighborhoods of the pixels to be processed in the image.The structural elements can be adaptively adjusted according to the features of the neighboring pixels.shape.The adaptive neighborhood mathematical morphology operator is defined and four basic operations are implemented.Compared with the classical mathematical morphology operation results,the adaptive neighborhood morphological operator proposed in this paper is objectively and objectively evaluated.Experimental results show that this paper The operator can retain more image information.Secondly,due to the wide application of mathematical morphology in image processing,the image processing speed has higher requirements,and the field programmable logic device FPGA is the mainstream large-scale programmable integrated circuit,which is to improve the mathematical morphology image processing.A good choice of speed,therefore,the proposed adaptive mathematical morphology operator combined with hardware FPGA to improve the mathematical morphology image processing speed,the binary image of the adaptive neighborhood mathematical morphology operator based on FPGA The processing system includes seven modules,namely,a control module,a data reading module,an expansion computing module,a corrosion computing module,an opening computing module,a closing computing module,and a template updating module,and applying Modelsim simulation software to complete adaptive mathematical morphology.The simulation experiments of four basic operations verify the logic correctness of each module and the feasibility of FPGA implementation of adaptive neighborhood mathematical morphology.Finally,the image processed by FPGA is compared with the image processed by MATLAB,which proves The correctness of adaptive neighborhood mathematics morphology based on FPGA.Finally,in order to evaluate the new operator proposed in this paper from an objective point of view,two evaluation indexes of structural similarity and edge preservation index are introduced,which are calculated for the results of classical mathematical morphology image processing and adaptive neighborhood mathematical morphology image processing.The index data was evaluated,and the image processing effect of the adaptive mathematical morphology algorithm proposed in this paper was qualitatively analyzed and evaluated by comparing the evaluation index data.At the same time,in order to make up for the lack of methods for quantitative analysis of mathematical morphology,a quantitative analysis method is proposed to calculate the pixel value of mathematical morphology image processing results,and the principle of adaptive mathematical morphology image is analyzed from the perspective of pixel value.And the effect of image processing.At the same time,the new operator is applied to the image edge extraction and filtering,and the edge extraction and filtering effects of classical mathematical morphology are compared.The advantages of the new operator in image processing applications are analyzed from the subjective and objective aspects.