Topic
Distance transform
About: Distance transform is a research topic. Over the lifetime, 2886 publications have been published within this topic receiving 59481 citations.
Papers published on a yearly basis
Papers
More filters
03 Dec 2010
TL;DR: A novel background modeling method that overcomes this limitation by relying on statistical models which use pixel phase instead of intensities to handle foreground detection under sudden illumination change.
Abstract: Effective foreground detection under sudden illumination change is an active research topic. However, most existing background subtraction approaches, which are intensity based, fail to handle this situation. In this paper, we propose a novel background modeling method that overcomes this limitation by relying on statistical models which use pixel phase instead of intensities. We first extract the phase feature of the pixel using Gabor filters. Then, a phase based background subtraction approach is proposed. In this approach, each phase feature is modeled independently by a mixture of Gaussian models and updated with a novel scheme. Since foreground pixels are scattered in the preliminary detection result, distance transform is implemented on the binary image which transforms the image into a distance map. We segment the distance image with a threshold and get the final result. Experiments on two challenging sequences demonstrate the effectiveness and robustness of our method.
13 citations
01 Jan 2004
TL;DR: A new automatic skeletonization algorithm for 3D vascular volumes is proposed that yields a more centered skeleton without complex post-processing and is also insensitive to boundary complexity and can be easily modified by the user.
Abstract: Extraction of the skeleton of vascular structures is an important procedure for computer aided analysis of vascular data. A new automatic skeletonization algorithm for 3D vascular volumes is proposed. Two types of distance maps and clusters, a set of connected points with the same property are used to represent the vascular structure. Using clusters representation, branch information can be retrieved efficiently. In each identified branch, preliminary points, defined as skeleton nodes, are derived hierarchically which are later interpolated to generate the skeleton. The algorithm was tested on MR angiography arterial and venous 3D vascular volumes. The extracted skeletons were reliable representation of the vascular structure. Compared to other 3D distance-based skeletonization algorithms, the new approach yields a more centered skeleton without complex post-processing. The skeleton is also insensitive to boundary complexity and can be easily modified by the user.
13 citations
Patent•
25 Jul 1997
TL;DR: In this article, the transparent degree is obtained in accordance with the distance of the object model to the image element unit, which is an obstacle to hide the character at the depth side by viewing from the visual point position of the pseudo camera.
Abstract: PROBLEM TO BE SOLVED: To obtain information of a character behind an object model without giving an unnatural feeling by executing display by which the transparent degree of respective image elements of the object model becomes higher as the distance of the object model to respective image element units as against the visual point position of a pseudo camera becomes nearer. SOLUTION: A transparent degree extracting means 12b is equipped, which sets the transparent degree of the respective image elements of the object model to higher degree as distance obtained by a distance extracting means 12a becomes nearer in accordance with the distance of the object model to the respective image element units as against the visual point position of the pseudo camera which is photographing the character and is movable by a moving means 32. Besides, the transparent degree is obtained in accordance with the distance of the object model to the image element unit, which is an obstacle to hide the character at the depth side by viewing from the visual point position of the pseudo camera. Image data of the image element unit of the object model is obtained through the use of the transparent degree and the image data is synthesized with image data of the image element unit of the character so as to display it in a display part.
13 citations
TL;DR: A propagation algorithm to obtain a skeleton with "good properties" based on the Euclidean distance map, which cleverly combines the centers of maximal balls included in the shape and the ridges of the distance map.
13 citations
Patent•
28 Dec 1996
TL;DR: In this paper, the authors proposed a solution to enable the control of parallax and the congestion distance with no delay in a comparatively short arithmetic time by determining any specified object out of plural objects.
Abstract: PROBLEM TO BE SOLVED: To enable the control of parallax and the control of congestion distance with no delay in a comparatively short arithmetic time by determining any specified object out of plural objects and controlling parallax so as to almost fix the parallax related to 1st and 2nd two-dimensional images concerning this specified object. SOLUTION: At a main body (game machine) 1 of three-dimensional image generator, a target object is determined as the specified object of concerned image based on a glance direction signal and concerning the determined target object, its distance Lobj is calculated by arithmetic. On the side of head mounted image display device HMD 2, based on the supplied object and the data of its background image, the display of image is performed by a three-dimensional image display part. Concerning the target object projected like this, offset is added to the display positions or the like of both left and right images so that a view distance can be equal with the congestion distance.
13 citations