Feature Analysis Of B-Ultrasonic Image Series

The edge feature extraction of images has always been one of the most important and the most basic research tasks in the field of computer vision. Contour tracking of motion objective has become one of the key subjects of computer vision. A lot of scholars have done a large amount of research in the fields these years, and have put forward various kinds of algorithms. However, in the fields of edge extraction and contour tracking for low SNR images, most existing algorithms become powerless.Herein, the idea is proposed that using Log Gabor wavelets whose phase information is preserved to correct the GVF Snake model. In this paper, b-ultrasonic image series, which are low SNR image series are dealt with. Though the result edge drawn by the deformable model will be influenced by a large amount of noises, the model has merits that traditional methods don't have. The principal merit is that the object dealt with by the model is a curve in succession. Thus a lot of mistakes caused by edge linking after edge detection are avoided, and moreover, GVF Snake has improved the general deformable model. It is able to enter concave shape area and has larger sphere of action. So, correcting GVF and making the model receive the correct result is the primary task of the paper.Here Log Gabor wavelets are used to reduce noises. Further, for the contradiction between noise reduction and keeping the vision impression, phase preserving and non-linear of Log Gabor wavelets to keep vision impression. From the above method, other results are obtained, that the restriction on σ and μ is reduced.The method put forward in this paper gives and produces a unified solution for extensive range vision problems. Finally in the edge extraction of high noise picture and the contour tracking of low SNR image series, this method has received relatively satisfied result.