Abstract: Difficulties faced in bone age automatic evaluation are bones position accuracy and extraction of bones interest area. Due to the influence of uneven illumination and the degree of development of irregular in X-ray bone images, the traditional method of image segmentation receives an unsatisfactory effect on extraction of bone segmentation. In order to achieve the accurate extraction of bone edges, a research method of extraction with AdaBoost cascade classifier based on ASM（Active Shape Model） algorithm is put forward, which enriches the application of automated bone age assessment system. According to the experimental results, the segmentation algorithm based on ASM can accurately locate X-ray images, making a good foundation for the next work.

Key words: bone age, AdaBoost cascade classifier, active shape model, hand bone extraction

摘要： 骨龄自动评估面临的困难是骨骼准确定位与骨骼兴趣区域提取。由于手骨X光图像存在光照不均及骨骼发育程度不规则等因素影响，传统的图像分割方法在骨骼上的分割效果不太理想；为了实现对手骨边缘的精确提取，结合AdaBoost级联分类器，提出基于ASM（主动形状模型）算法的手骨边缘提取方法，丰富了骨龄自动评价系统的应用研究。实验表明，基于ASM算法的手骨分割能有效对手骨X射线图像进行准确的定位，为骨龄自动化评价系统的下一步工作奠定基础。

关键词: 骨龄, AdaBoost级联分类器, 主动形状模型, 手骨提取