Auto-segmentation of cartilage and bone areas in thoracic vertebrae images

The automated segmentation of the cartilage and bone areas in the thoracic vertebrae images based on OHTA color systems was discussed in this paper. The thoracic vertebrae images were captured with digital camera and were saved in computer. The image in RGB format was first transformed into OHTA color system; then I₂ and I₃ values of OHTA system were effective in segmenting cartilage and bone areas, respectively. After applying binary image, morphological erodes, hole filling and area threshold on I₂ and I₃ images, the geometric properties of cartilage and bone were obtained. The results of this work contribute to the automatic computer vision system for beef maturity detecting and grading.