Segmentation and cavity filling of color image from stained trabecular sections based on HSV space / 生物医学工程学杂志

ABSTRACT

A approach to segment the color image based on Hue-Saturation-Value (HSV) space was proposed, and it was used to segment the color image digitized from the bone sections stained with Masson-Goldner's Trichrome by CCD camera. According to HSV approach, color image was transformed from RGB space to HSV space at first, and then the image was segmented by using the threshold value of hue (90 < H < 150) and saturation (S > 0.25) to find the image of the green color stained trabecular. And then used the threshold value of saturation (S < 0.2) and value (V > V(background) x 0.95) to find the unstained high brightness field. At last, unstained high brightness fields which were concluded in stained trabecular image were filled with the color of trabecular, so the completely trabecular image could be drawn. The results showed that HSV approach was fast and simple, and it could be an efficient automatic algorithm.