This paper proposes a segmentation method and a three-dimensional (3-D) volume calculation method of cysts in kidney from a number of computer tomography (CT) slice images. The input CT slice images contain both sides of kidneys. There are two segmentation steps used in the proposed method: kidney segmentation and cyst segmentation. For kidney segmentation, kidney regions are segmented from CT slice images by using a graph-cut method that is applied to the middle slice of input CT slice images. Then, the same method is used for the remaining CT slice images. In cyst segmentation, cyst regions are segmented from the kidney regions by using fuzzy C-means clustering and level-set methods that can reduce noise of non-cyst regions. For 3-D volume calculation, cyst volume calculation and 3-D volume visualization are used. In cyst volume calculation, the area of cyst in each CT slice image equals to the number of pixels in the cyst regions multiplied by spatial density of CT slice images, and then the volume of cysts is calculated by multiplying the cyst area and thickness (interval) of CT slice images. In 3-D volume visualization, a 3-D visualization technique is used to show the distribution of cysts in kidneys by using the result of cyst volume calculation. The total 3-D volume is the sum of the calculated cyst volume in each CT slice image. Experimental results show a good performance of 3-D volume calculation. The proposed cyst segmentation and 3-D volume calculation methods can provide practical supports to surgery options and medical practice to medical students.

No abstract

Computer vision algorithms using depth and color data (RGBD image) acquired at the same time are gradually getting an attention. Depth data is useful three-dimensional (3-D) information. However, the depth data has a problem that it includes depth hole regions. In this paper, a depth hole filling method is proposed, which is suitable for 3-D reconstruction, with the detail of depth region of an object preserved. Keywords-depth hole; 3-D reconstruction; RGBD image; depth map enhancement I.INTRODUCTION Three-dimensional (3-D) information is obtained by using a range camera. The early type of range camera is composed of radio frequency (RF) modulated light sources and the receiver that includes phase detector [1]. To measure the distance between an object and a camera, this type of range camera emits RF-modulated light. Then, the receiver of a range camera measures the distance by detecting phase difference of RFmodulated light. Another type of range camera consists of structured infrared light (SIL) source and infrared camera [1]. SIL source emits regular patterns of infrared light into scene, then infrared camera produces image by detecting emitted patterns of infrared light. The distance between a range camera and an object is measured by observing patterns in the image. However, the depth image acquired by a range camera contains regions that do not have any depth information, which is called the depth hole problem. To solve the depth hole problem, various algorithms have been presented [2,3]. To fill holes, Yang et al.'s method uses the distribution of the depth values around depth holes [2]. The value that fills depth holes is computed by averaging peak values of the distribution of the depth values around depth holes. After that, depth map refinement proceeds by using the color image. Xu et al.'s method uses the maximum depth value in a window of a certain size to fill depth holes [3].Recently, visible-light camera and range camera are combined to obtain color and depth information at the same time [4]. This type of camera is called RGBD camera, and the acquired information is called RGBD image. RGBD image is very useful information because it represents not only color information but also 3-D information. However, RGBD image has the depth hole problem, which degrades the performance. For example, the depth hole problem is a critical factor in 3-D reconstruction. In [5], Lim et al. use scale invariant feature transform (SIFT) features to match multiple RGBD images acquired at different viewpoints. Most of SIFT features are located at edges of RGB image. The depth holes also lie at edges of RGB image. Thus, most of SIFT features do not have