Image segmentation is one of the classic problems in the computervision field. Although a lot of successful operators and algorithms have been proposed, fuzzy image segmentation does not always achieve satisfactory r...
详细信息
Image segmentation is one of the classic problems in the computervision field. Although a lot of successful operators and algorithms have been proposed, fuzzy image segmentation does not always achieve satisfactory results. This paper is inspired by Positive Selection Algorithm and Negative Selection Algorithm and, is based on the mechanism and process where T-cell is activated by the MHC molecule. A new positive selection algorithm is introduced which establishes so-called templates set for immune detection. This algorithm is based on processing of image information represented as a gray value statistic rather than arithmetic gradient formulation. It is comprised of a template set not just a single template. Therefore it gives good results for different images. The presented algorithm is used for image segmentation into objects, background and fuzzy edge in fuzzy infrared images.
A novel rapid large noise denoising method for range image is *** a 2D scan line acquired by a laser range finder,distribution features of large noises(LNs)are analyzed,and then a mathematical representation of the fe...
详细信息
A novel rapid large noise denoising method for range image is *** a 2D scan line acquired by a laser range finder,distribution features of large noises(LNs)are analyzed,and then a mathematical representation of the features is provided by defining a few new coefficients. Subsequently,a rule-based distinguishing criterion is formulated to detect two types of LNs:dropouts and invaders. The traditional mean filter is improved by an automotive non- LN neighbor searching procedure.A compositive algorithm with a very low computational complexity has been implemented as an embedded module in our self-developed software with *** experiment on real range image is performed,and the results indicate that the proposed method can detect all the large noises accurately and denoise them with a significant *** is proven that the method is suitable for practical applications on industrial or other fields.
During the last years, the need for security-oriented surveillance systems has grown higher and higher. Nowadays many public environments, such as airports, train stations, etc. are monitored by some sort of video-sur...
详细信息
During the last years, the need for security-oriented surveillance systems has grown higher and higher. Nowadays many public environments, such as airports, train stations, etc. are monitored by some sort of video-surveillance system in order to detect or prevent security issues. The involved technology ranges from the use of plain closed-circuit cameras (CCTV) to sophisticated computer-based video processing systems. The CCTV approach has been the only feasible choice in the past, and it is still widely used, however its limits are more and more evident: the increase of the number of sensors (modern surveillance systems can use hundreds of cameras) is often not matched by an adequate number of human operators, whose attention is spread on many different tasks and quickly decreases over time. Modern computer-based systems try to face these problems using automatic video analysis and understanding techniques, in order to cover wide areas and simultaneously highlight only the potential security issues and thus requiring the attention of a human operator only in a limited number of cases (e.g. [6, 5]). The research in this field has been very active and produced many techniques for video analysis and interpretation, but many works are limited to the use of static cameras. Only recently the research community started focusing on more sophisticated sensors like Pan-Tilt-Zoom (PTZ) cameras, and the research on dynamic, active networks of PTZ cameras is still limited (for an example of some recent works in this field, see [1]). Many of these works focus on exploiting the dynamic features of a network of PTZ cameras to improve tracking performance [3, 4, 13, 10, 12], while relatively few works address the problem of optimizing the camera coverage of the monitored area according to specific criteria. Angella et al. [2] propose a method to maximize the area coverage by using a 3D model of the observed zone, but their work only aims at finding a good initial camera displacemen
Automated scene interpretation has benefited from advances in machine learning, and restricted tasks, such as face detection, have been solved with sufficient accuracy for restricted settings. However, the performance...
详细信息
Automated scene interpretation has benefited from advances in machine learning, and restricted tasks, such as face detection, have been solved with sufficient accuracy for restricted settings. However, the performance of machines in providing rich semantic descriptions of natural scenes from digital images remains highly limited and hugely inferior to that of humans. Here we quantify this "semantic gap" in a particular setting: We compare the efficiency of human and machine learning in assigning an image to one of two categories determined by the spatial arrangement of constituent parts. The images are not real, but the category-defining rules reflect the compositional structure of real images and the type of "reasoning" that appears to be necessary for semantic parsing. Experiments demonstrate that human subjects grasp the separating principles from a handful of examples, whereas the error rates of computer programs fluctuate wildly and remain far behind that of humans even after exposure to thousands of examples. These observations lend support to current trends in computervision such as integrating machine learning with parts-based modeling.
Image matching is a fundamental task for computervision. However, due to the local ambiguities induced by repetitive patterns in most of man-made objects or scenes, matching images of repetitive patterns is still a c...
详细信息
Image matching is a fundamental task for computervision. However, due to the local ambiguities induced by repetitive patterns in most of man-made objects or scenes, matching images of repetitive patterns is still a challenging problem even the viewpoint changes of them are not very large. This paper is primarily focused on the problem of matching images containing repetitive patterns and proposes a novel method based on pairs of interest points to solve the problem. It starts from matching pairs of interest points and then obtaining point correspondences from the matched point-pairs based on the low distortion constraint, which is meant that the distortions of point groups should be small across images. By combing pairs of interest points, local ambiguities induced by repetitive patterns can be reduced to some extent since information in a much larger region is used. Moreover, owing to our newly defined compatibility measure between one correspondence and a set of point correspondences, the obtained point correspondences are very reliable. Experimental results have demonstrated the effectiveness of the proposed method as well as its superiority to the existing methods.
暂无评论