最新刊期
卷 4 ， 期 12 ， 1999
- 摘要：Many feature extraction methods begin with edge detection. Conventional color edge detection usually uses concepts similar to monochrome images. Consequently much color information are not thoroughly utilized. The idea of applying vector order statistics to edge detectors has been proposed to solve this problem. Because real world images contain distinct edges at various resolutions, effective extraction may require the combination of edges across several scales. Several researchers have proposed methods that employ filters at multiple scales. According to these approaches, we propose a scheme for multi scale color edge detection based on vector order Prewitt operators. Each channel of the color image is initially convolved with two Gaussian smoothing functions, one scale of that is higher and the other is lower. Then the vector order Prewitt operators are applied to the two smoothed images. Furthermore, we derive a rule based on the edge behavior in scale space to combine the two scale edge information. At last, the locations of the edge are detected after thinning and thresholding. Experiments have shown that the edges recovered by the proposed method are more integrated and accurate than those obtained by the vector order Prewitt operators running only at one scale. In addition, the edges detected by vector operators are much better than those detected using scale operators. It is indicated that using vector order statistics and multi scale approach is effective in color image edge detection.关键词：Color image processing;edge detection;Multi scale;Regularization2257|226|0更新时间：2024-05-08
- 摘要：The disadvantage of PSNR, the most common objective measure to evaluate the decompressed images, is pointed out. A new measure named DSNR (Detail SignalNoise Ratio) is therefore proposed. DSNR can be calculated by abstracting energy of detail signal and noise. Using DSNR, the quality of decompressed images can be directly measured without the original (uncompressed) images. Our experiments show that DSNR of different images in the same scene changes congruously with PSNR, thus can be used to evaluate the quality of decompressed images and image sequences.2612|123|0更新时间：2024-05-08
- 摘要：Image morphing method, which is based on the pixels position and color interpolation between two images, has been widely used in image interpolation. The former methods can't keep 3D shape, even though 2D shape. A new morphing method adopted viewpoint morphing technique. The 3D shape of the object can be kept if the projection transformation (called camera matrix) from 3D to 2D can be predetermined. It is difficult to gain camera matrix from the two images by the existing methods and techniques. This drawback confines to the applications of the view morphing. In this paper, a new morphing method is presented. We proved that the interpolatory view is correspondent to a viewpoint. The method needn't predetermine the camera matrix. And it is of simple computation and easy realization.2151|184|0更新时间：2024-05-08
- 摘要：Pattern recognition is derived from the will to imitate the sense of the creatures, object searching in an image aims at imitating the human' s visual ability. In this paper, utilizing the results of the vision neurophysiology and the psychology of cognizance, and aiming at the specific pattern location in the natural environment, we put forward a Bionic Object Exploring Strategy——BOES. In a license plate recognition system, an algorithm of succinctness and practicality on this strategy has developed to solve the difficulty of locating license plates rapidly and precisely. In our algorithm, to deal with the complex, uncontrollable background of the natural environment and the variability of the color, brightness and size, we extract the license plate pattern' s most permanent characters, design a set of filters to mask the background, organize the image data in a series of steps by spatial and frequencial filtering, and divide the task into several layers. This system can locate the license plate in an image of complex environment within 0.5 second on a PentiumⅡ 300Hz machine, and the rate of success is over 95%.关键词：Pattern location;BOES strategy;PPC character;License plate location2136|139|0更新时间：2024-05-08
- 摘要：An improved Dijkstra algorithm is set forward on the basis of priority queue based on quad-heap. It is proved in this paper that quad-heap is the optimum amongk-ary heap. The presented algorithm decreases the complexity of conventional Dijkstra algorithm toO(logn). Considering the dynamic characters of transportation applications, an inverse running sequence is presented based on the inverse adjacent list data structure. It is more suitable for the dynamic applications of GIS-T.2341|132|0更新时间：2024-05-08
- 摘要：This paper is a review on multimedia researches and applications in China, 1998. Papers about multimedia technology are distributed on various journals. We saw about 2220 papers published on 12 Chinese journals in 1998, from which we selected 334 ones on multimedia technology and made analyses on them. We present here an overview on the progress in multimedia technology in China, 1998. This will be convenient for researchers looking up reference, and helpful for editors compiling journals and for authors contributing papers.2152|144|0更新时间：2024-05-08
- 摘要：This paper describes the Hyperspectral Image Processing and Analysis System (HIPAS). The HIPAS, built on Interactive Data Language (IDL), and implemented on Windows NT workstations, meets the requirements for rapid preprocessing of imaging spectrometer data and easy prototyping of algorithms. Integrated with a spectral library, which was implemented on the FoxPro, a popular used database environment in Windows NT platform, the spectral analysis model was established to support the hyperspectral image analysis. The design, structure and main characteristics of HIPAS are described in this paper. The functions needed in the hyperspectral image processing system and the applications of object oriented design method used in the design of software system are discussed. Designed using the object oriented methods, the HIPAS system is robust and easy to be expanded. It shows that based on the personal computer hardware system and using an optimized software algorithm, the basic requirement of hyperspectral image processing can be satisfied.3543|156|0更新时间：2024-05-08
- 摘要：In this paper, an approach based on wavelet decomposition is developed for image segmentation. After making the image two dimensional wavelet transform, fractal dimension and texture features of subchannels are extracted for texture segmentation. The experimental results also demonstrated its satisfactory efficency both for simulated textured image and multispectral remotely sensed image.关键词：Image segmentation;Texture;Wavelet decomposition;Fractal;Feature extraction2136|111|0更新时间：2024-05-08
- 摘要：This paper aims at the random point,line and surface in vector GIS, based on the analysis of the point's probability density function, image and the error interval, error ellipse which are induced by positional uncertainty, elaborates the random line's, surface's probability density functions and images by means of the stochastic process and random field theory, presents the proper reasons of the line's error band and surface's error donut. The visual figures according to different conditions are given.2041|103|0更新时间：2024-05-08
- 摘要：The development and problems as well as perspective of optical neural networks (ONN) are discussed in this reviewing paper, along with the presentation of two typical 2-D hybrid ONN models interconnected with lenslet array and hologram elements respectively. Novel algorithms and topologies of ONN are also introduced.关键词：Optical neural network;neural network;Optical interconnection;Optical computing3613|113|0更新时间：2024-05-08
- Postal code：100190
- Tel：010-58887035/58887030/58887418 Email：jig@aircas.ac.cn
- Technical support is provided by Beijing Founder electronics co., LTD 京ICP备05080539号-4
京公网安备11010802024621 - It is recommended to read the content of this site in Chrome&IE9+. Please switch to extreme mode in browser 360.
- Cookies We use cookies to help provide and enhance our service and tailor content. By continuing, you agree to the use of cookies.
0