Research Of Interactive Image Segmentation

Computer vision is an important branch of image signal processing, it is multi-disciplinary in that it involves multiple areas such as signal processing, computer, mathematics and psychology, and it has important applications in military, manufacturing, medical industries as well as commercial image and graphics software. Interactive image processing was developed in recent years, and has been receiving more and more attention. On the base of concluding interactive segmentation algorithm, some approaches in interactive segmentation are researched. In this dissertation, we attempt to have an in-depth investigation on the Live Wire algorithm and random walk algorithm. The main work is as follows:Considering the sensitive to noise, can not correctly distinguish between the strength image edge and the weak image edge and the speed disadvantage of the traditional Live Wire algorithm, a novel Live Wire algorithm based on Pusle Coupling Neural Network (PCNN) is proposed. PCNN is used to obtain the shortest path between the two points by user. We achieved image segmentation. The experimental results show that the improved algorithm is feasible and credible, and gives an ideal inage segmentation outcome, The speed of operation is increased by proposed method, compared with traditional methods, it is more fast and steady. Especially when dealing with a larger image, the rate reflected a clear advantage.Againsting the traditional random walk algorithm description of a single image information and target profile susceptible to background interference problem, this paper improved random walk image segmentation algorithm. We establish a weight function based on texture similarity of expression. Through using of the Gabor filters, texture features is introduced into the random walk algorithm in order to highlight the image structural information.In the end, Through the Dirichlet boundary conditions, we achieved image segmentation. The experimental results show that the improved algorithm describes the image structural information better. Compared with traditional methods, the improved algorithm reflected better applicability and segmentation accuracy in the image segmentation details.