Research On Edge Detection Algorithms Of Gray Image

In the information age,people's demand for information acquisition is increasing day by day.Especially,the demand for useful information from images becomes very important in all neighborhoods.Edge information acquisition of objects in images is one of the important research directions of image information acquisition,and is also the basis of image analysis.This paper focuses on the edge detection of objects in gray images,and improves the existing edge detection algorithms.The main work is as follows:Firstly,an improved edge detection algorithm for multi-directional structural elements is proposed.Based on the introduction of morphological edge detection operator,the influence of multi-directional structural elements on edge detection results is analyzed,and the traditional multi-directional structural elements edge detection algorithm is improved.The basic idea of the improvement is to change the traditional method of calculating the fixed weight of detection results corresponding to multi-directional structural elements from the original method of calculating the average value of all detection results to calculating the ratio of edge variance of detection images on eight different directional structural elements.The simulation results show that the improved method is more accurate than the traditional method.Secondly,an improved edge detection algorithm for multi-scale structural elements is proposed.Based on the introduction of structural element selection and multi-scale edge detection algorithm,an improved multi-scale structural element edge detection algorithm is proposed,which regards structural elements of different scales as a composite structural element.The algorithm can detect the edge of the image at different scales and retain the details of the edge to the maximum.The simulation results show that the improved method can detect the edge details more accurately.Thirdly,a multi-scale and multi-structure edge detection algorithm based on morphology is proposed.Traditional multi-scale morphological detection algorithm and multi-structure morphological detection algorithm use mean or fixed weights to determine the proportion of final detection results.They do not connect each atomic detection result with the original image.When processing different images,the results of detection differ greatly and do not have universality.Aiming at this defect,a multi-scale multi-node with adaptive weights is proposed.Method of structural morphology detection.The basic idea of the algorithm is that the improved multi-scale detection results proposed in the first work can detect more accurate edge details,and the improved multi-structure detection results proposed in the second work can detect more abundant edge details.The weighted averaging of the two results can not only improve the accuracy of edge characterization,but also enhance the richness of edge information.The simulation results show that the improved edge detection method has greatly improved the quality and quantity of the detection results.Fourthly,a fast edge detection algorithm based on fuzzy theory is proposed.Aiming at the high computational complexity of Pal King's fuzzy edge detection algorithm,which uses power function membership function,and the disadvantage of losing low gray value edge information easily,an improved fast edge detection algorithm is proposed.The basic idea of the algorithm is to use a simple quadratic membership function to reduce the computational complexity of the algorithm,and to reduce the number of iterations of the conversion function so as to reduce the loss of edge information.The simulation results show that the improved algorithm greatly reduces the computational complexity on the basis of slightly increasing the error rate.