Quantitative evaluation of repairing effect of bone grafting material is one of the essential studying subjects. However, traditional evaluation methods are subjective and qualitative. In this paper, the region of new bone from a bone repairing biomaterial planted image is extracted based on color image segmentation and then statistically analyzed to evaluate the property of bone grafting material quantitatively. HSI color model, which corresponds with people’s vision system for color is used to achieve ideal segmental results and effective utilization of color information. The S and I variable are used as thresholding condition for image segmentation, thereby obtaining the area of new bone. The effect of BMP contained in BMP/α-TCP is estimated furthermore. Experimental results show that the composite BMP/α-TCP induce more bone than pure α-TCP in virtue of BMP. This study provides theoretical and experimental suggestions for clinical applications of BMP/α-TCP.