基于邊緣檢測(cè)的圖像分割方法研究-畢業(yè)論文

1、本科畢業(yè)論文 本科畢業(yè)論文(科研訓(xùn)練、畢業(yè)設(shè)計(jì))題 目：基于邊緣檢測(cè)的圖像分割 目：基于邊緣檢測(cè)的圖像分割方法研究 方法研究姓 名：學(xué) 院：軟件學(xué)院系：專 業(yè)：軟件工程年 級(jí)：學(xué) 號(hào)：指導(dǎo)教師（校內(nèi)）： 職稱： 年 月IIAbstractA central problem, called image segmentation, is to distinguish objects fr

2、om background. For intensity images (ie, those represented by point-wise intensity levels) four popular approaches are: threshold techniques, edge-based methods, region-based techniques, and connectivity-preserving relax

3、ation methods.Edge detection is an important part of image segmentation. Currently edge detection methods include classical algorithms like edge detection derivative and second-order differential edge detection, and the

4、new methods based on some Modern technology. Then a detailed description of several algorithms is given, which include gradient algorithm, the algorithm Roberts, Sobel algorithm, Prewitt algorithm, the algorithm Kirsch,

5、Laplacian algorithm, Gaussian Laplacian algorithm (Laplacian of Gaussian) and Canny algorithm. And the operator of the various edge detection effect is on show throught the experimental results of a VC++.NET program. Aga

6、inst Canny algorithm, an adaptive multi-scale Canny edge detection method is given. Adaptive multiscale Canny edge detection algorithm can be anti-noise performance and the details to maintain the balance of good edge. C

7、omparing to the result of Canny edge detection, the edge that adaptive multiscale Canny edge detection algorithm detects is relatively clear, with high signal-to-noise ratio. The final article briefly introduces the mode

8、rn edge detection methods, including those based on wavelet edge detection, based on morphological edge detection, based on fractal theory of edge detection, based on fuzzy edge detection, Based on artificial intelligenc