In Skeletal Age Identification System, Positioning Research, Identify With The Radial

With the development of Computer Technology, Digital Image Processing is one of hot research fields in recent years, its development and wide application enhance the level of automation. Assessment of skeletal maturity from a man' s left-hand wrist radiograph is a new technology in the recent thirty years and being developed. And now, It is crucial how combine digital image processing technique with assessment of skeletal maturity to come into being Computer-assisted Bone Age Assessment.Bone age assessment is based on an analysis of ossification centers in the carpal bones and epiphyses of tubular bones including distal, middle, and proximal phalanges as well as radius and ulna. Traditionally, bone age assessment was made manually by comparing the radiograph with textual and pictorial information from a standard alts. This is a time-consuming and tedious process, and the result is different each other. By using computer vision techniques, it is quite practical for us to develop a computerized analysis system and automated the skeletal system estimating procedure, it also can conquer these difference.However, as far as the computer-assisted bone age assessment concerned, X-ray image analysis plays an important role in this system, and how to recognize skeletal maturity is also very important by using digital image processing technique to extract correlative features.This paper does more detail research in the area af position and feature extraction. Study conventional image processing algorithms to amalgamate and improve convenient for the following procedures. Based on the knowledge we already know from a X-ray image , we use the Thinning Tracing Sobel edge arithmetic operator and projection algorithm to realize position. It is effective to apply ina X-ray palm segmentation. Although a great number image edge arithmetic operators exist, results show that an optimal method for a correct edge extraction has not yet found. We developed the Laplacian of Gtiass edge arithmetic operator, and realize the radius recognition based on fuzzy pattern matching. Finally , we introduce the image segmentation methods in order to show the improving direction of this paper.All of the algorithms discussed in this article have performed in the software development platform of Microsoft Visual C++6.0. The result is useful but we still have a lot work to do.