一种基于角点检测方法的骨龄图像关键点定位

A location method for the key points of the skeletal images based on corner detection

在骨龄自动化评定系统的研究中,指骨关键点的定位是其中的一个比较重要的环节,对后续工作研究关系重大。由于指骨关键点是手掌轮廓变化最大的地方,因此把图像中常用的角点检测方法引入到医学图像中,在分析常见的角点检测算法的原理与不足的基础上,提出一种改进的算法来定位指骨关键点。通过实验,并与其它同类方法比较,结果表明,在指骨关键点的精确和快速定位方面,该方法都达到了比较好的效果,可直接应用到骨龄自动评价中。

In the research of the automatic bone age assessment,the key points location for the phalanx is the one of key factors,which determines mostly the following research.The key points of phalanx are the most bendest place around the palm.In the article,a common corner detecthon metod is introducted into the medical images.Based on analyzing the principles and shotcomings of the common corner detection algorithms,an improved algorithm is proposed,which is used for the location of the phalangeal key points.Through a large number of experiments and comparisons with other similar methods,the result of an experiment showed that the proposed method is getting a better results in the accurate and fast location,which can be directly applied in the assessment of the bone age.