314 Calculation of Injured Area in Forensic Image Based on Segmentations

For an accurate evaluation of the injured area in forensic medical image, four image segmentation methods including iteration threshold segmentation, Otsu method, Local threshold segmentation, K-means clustering segmentation are presented. The comparison of four image segmentation methods to get experimental data is necessary. The numerical measurements are required to tag the injured region by segmentations to partition the foreground and background and the mathematical morphological processing is used to smooth the edges and eliminate the holes and spots in the image. Then the calculation of the area by counting the pixels of injured region by scanning of the image is proposed. Therefore, the effective procedure of the area measurement for arbitrary shapes in forensic medical image is given. The result shows that four methods are effective to segment the forensic medical image and the three errors in the injured area of them are less than 0.5%.