Study On Multi-Objective Boundary Method In Static Image

Image information processing occupies an important position in real life.Contour extraction is a kind of image information processing technology.It is one of the research contents in the field of image processing.It is widely used in forest resources and environment monitoring and other remote sensing information data analysis,road traffic Situation control,medical image analysis,and license plate,fingerprint identification,etc.There is a lot of research on the extraction of image boundary data using the boundary tracking method,but most of the existing boundary tracking methods are based on binary image processing,and binary image conversion is required for grayscale images.Due to noise and other problems in grayscale images,the converted binary image is often not ideal.The algorithm does not have a good choice for the starting point.It takes a long time for each algorithm to track the screening and select the starting point.These tracking methods are prone to leak tracking when tracking multiple targets.As for these defects,a multi-target boundary tracking method based on static grayscale images is proposed in this paper,and has a higher requirement on the starting point,and proposes the definition of seed points.In the tracking of some medical images and noise grayscale images,this method has a good edge tracking effect.From the analysis of noise images,this method has a certain degree of noise immunity.We divide this paper into following four parts:In the first part,according to the geometric characteristics of the ridge formed by the boundary points in the filtered gradient map,a new seed point definition is proposed.From the seed point,all objectives can be tracked.We also can ensure that all seed points are located on the objective boundary,and there is at least one seed point on each target boundary line.In the second part,a multi-target boundary tracking method based on static grayscale image is proposed.It mainly aims at the multi-objective grayscale image boundary tracking and solves some defects in the previous methods.The third part,we propose a stop mechanism to deal with images with non-closed boundaries.In the last part,the part of experiments.We have designed a number of theories proposed in the third part of numerical experiments.The tracking of noise images with different degrees shows that the method has a certain degree of noise immunity.By contrast with the Eight-neighborhood tracking method and the Canny method,the boundary of the method tracking is better.The boundary integrity is higher and there are no redundant boundaries.At the same time,the effects of the two parameters in the algorithm on the experimental results are also analyzed in detail.Experimental results show that the proposed method can adaptively track the boundaries of multiple targets in complex grayscale images,obtain better tracking results,and have a certain degree of noise immunity and strong robustness.