Research On Image Segmentation Methods Of Ochotona Curzoniae Based On CV Model

Ochotona curzoniae is an endemic species and key species in the alpine meadow of the Tibetan Plateau and also it is a main kind of organism that destroyed the grassland ecology of the Tibetan plateau.In order to prevent the dangers of the Ochotona curzoniae,we need to study the living habits of Ochotona curzoniae and investigate the degree of harm of Ochotona curzoniae.With the development of sensing technology and image processing,we can provide an objective basis through intelligent monitoring system to control the damage of Ochotona curzoniae.The object detection of Ochotona curzoniae is a key technology in the intelligent monitoring equipment because it can provide the object contours feature for behavior analysis of Ochotona curzoniae.The object detection of Ochotona curzoniae is very difficult,because the Ochotona curzoniae video possess the characteristics of complex background,low contrast,the object color with intensity inhomogeneity,diversity and mutability.In this paper,the Chan-Vese model is used as the basic model,and the Chan-Vese model is improved according to the image characteristics of the Ochotona curzoniae,so as to realize the image segmentation of the Ochotona curzoniae.As the internal fitting value of Chan-Vese model could not fully represent the object area information,the K-means clustering method is proposed to the pixels in the initial contour of the image.The internal fitting value is obtained by linear combing of intensity information of the image after filtering and the cluster center point value.It can improve the adaptability of the model to the complex object,so as to obtain the complete object area information.In addition,rectangular Dirac function was used to replace regularized Dirac function in the energy function of Chan-Vese model,and the calculation of level set evolution equation could be limited to the zero level set so as to avoid the influence of the image background disturbance on the segmentation result,in the same time,the computation of image segmentation is reduced.The improved Chan-Vese model can accurately and quickly segment the complex object image.Aiming at the characteristics of Ochotona curzoniae sports mutation,this paper presents a fast object detect method based on Spatio-temporal domain.In the time domain,the moving object location and rough segmentation image are acquired through the background subtraction method,which provides initial contour for spatial domain segmentation and reduces the processing area of spatial domain.In the space domain,using the improved Chan-Vese model which presented in this paper to segment the rough segmentation image,then the accurately object contour is accurately obtained.The experimental results show that the time domain processing method is not affected by the characteristics of the object's movement behavior,and the position of the Ochotona curzoniae can be accurately located.The spatial processing method not only can quickly and accurately extract the contour of the Ochotona curzoniae,but also avoid the need for the Chan-Vese model to interactively set the initial contour information problem.This method can be continuously and accurately segment the image sequence of the short-term Ochotona curzoniae.