基于OpenCV的X光手指骨图像分割方法

X-ray Finger Bone Image Segmentation Method Based on OpenCV

骨龄评价在预防医学、临床医学、体育科学和司法领域等都有着广泛的应用。随着计算机技术、图像处理技术的快速发展,基于图像处理的骨龄识别系统是目前骨龄评价的发展趋势。在图像处理中,图像预处理和图像分割技术占有很大的比重,分割质量的好坏对后期的图像识别影响很大。目前图像的分割方法有多种,文中针对骨龄测评系统中的X光片手指骨,按照CHN标准,14块特征骨块中,取其中8块进行研究,对除了拇指外的四根手指进行分割识别,使用基于Haar分类器的目标检测,来完成对手骨的识别、分割。识别出手指后,记录的实际上是手指边缘的坐标信息,依据坐标对手指进行分割提取,为后续的特征点标记以及特征值的计算打下基础。实验结果表明,使用Haar分类器进行手指的识别提取,速度快,准确率较高。

Bone age assessment plays an important role in preventive medicine, clinical medicine, sports science and judicial field. With the rapid development of computer technology and image processing technology, image processing -based bone age recognizing system becomes the tendency for bone age assessment. In image processing, image preprocessing and image segmentation technology occupy a large proportion. Segmentation quality has a great influence on later image recognition. Currently there exist a variety of image segmentation methods. In this paper,in view of the bone age assessment system of X-ray finger bone,with the standard of CHN,8 out of 14 featured bones are chosen to study. One except the thumb is chosen to identify the division of the four finger bones, target detection based on Haar classifier is used to complete the recognition and segmentation of the bone. After the recognition, what is recorded is the coordinating information of finger edges. Segmentation of fingers clone on coordinate lays foundation for later marking of featured dots and calculating of feature value. Experimental results show that using Haar classifier for extracting finger recognition has high speed and higher accuracy.