Research On The Representation Of Image Boundary And Its Application In The Design Of Paper Cut-out

With the wide application of image technology in every field of society, more and more experts and scholars pay attention to the development of image processing. Boundary is a small part of image and is also an important feature of image. Its technology can be applied to image recognition, industrial design, 3-D reconstruction in medical image, generation of virtual scene. Boundary processing includes boundary extraction and boundary representation. It is a basic research in image processing and analysis and is also a problem which isn't complete solved. Image processing is an application-oriented subject and most of its theories and technologies are closely related to practice. There are almost no general theories and technologies of image processing, so does boundary processing. Research on the technology of boundary processing for practice can not only solve the problem, but also enrich the technology of image processing. Therefore, the research of boundary processing is an important and significant work.In this paper, we first introduce the development of image processing. Then we analyze some new methods of image processing and the current domestic and international research status on image boundary. To overcome these limitations, we improve and propose some algorithms. Theoretical analysis or experimental results have proved that the proposed methods are feasible and effective. The main work in the thesis are as follows:(1) Research on the representation of image boundary using chain code.Chain code is a common boundary-representation of image. The representation of image boundary using chain code is researched first. To overcome the limitation of the chain code, an improved method of chain code is proposed. The dualistic group is proposed to replace coding one by one. It saves storage effectively, which uses the high bits and low bits in a byte to access the two items in the dualistic group respectively. Such work corresponds with chapter 2.(2) A new boundary extraction algorithm in a binary image.The boundary of binary image extracted by traditional methods is not one pixel width. In order to overcome this limitation, we analyze the feature of the boundary line. Then we induce the rule of boundary pixel and propose an operator to extract the boundary. Using this operator, the algorithm can directly extract the boundary pixels from the binary image and get one pixel width boundary. This part corresponds to chapter 3.(3) A control point based algorithm of vectorization in a binary image.In chapter 4, the vectorization of binary image has been researched. We propose a method of vectorization to convert the image into the graphic. First, the boundary is extracted by tracing. Then, the corners on the boundary line are detected by the SUSAN approach and the control points between each two neighbor corners are determined by the offset distance. Last, the cubic B-Spline curve is used to fit the boundary.