最新刊期
卷 10 ， 期 2 ， 2005
- 摘要：Speckle, appearing in synthetic aperture radar (SAR) images as granular noise, is generated by the coherent processing of radar signals. It severely affects the image qualities. To alleviate deleterious effects of speckle, various ways have been devised to suppression it. An ideal algorithm should smooth the speckle without blurring edges and fine detail. But most algorithms cannot satisfy these two demands very good.This paper investigates the problem of speckle suppression comprehensively and thoroughly. It first analyzes the principle of speckle forming, then describes the general approaches of speckle reduction filtering, discusses several well known speckle reduction filters, compares the algorithms qualitatively. To compare speckle smoothing algorithm quantitatively, a set of performance criteria is established. In the end, conclusions are made and the possible future work is explained.3253|204|0更新时间：2024-05-07
- 摘要：The detection of regions of interest (ROI) introduces visual attention mechanism from the human vision to the image analysis. It is significant to improve the efficiency of the existing image analysis system. In this paper, the current technique for detecting ROI in a static image is surveyed. After the background of ROI detection is introduced, the mechanism of visual attention in human visual system is addressed firstly. And then the main algorithms of ROI detection are classified into two categories including bottom up method and top down method. Each of them is discussed in detail respectively. The primary applications of ROI are described afterward. At the end, several research trends of ROI detection are given.3435|220|0更新时间：2024-05-07
- 摘要：At low bit rate, the images encoded by standard wavelet coders inevitably undergo the so called ringing phenomena near sharp edges. A low bit rate image compression algorithm based on edge adaptive wavelet transform is proposed. At the encoder side, firstly the strong edges of an image are detected and encoded as additional information; Then the wavelet transform is carried out in such a way that no filtering over previously detected edges is performed; Finally, the wavelet coefficients are encoded by the wel known EBCOT algorithm. Especially, a new method to overcome the boundary effects in wavelet transform of segmented data is proposed. Compression examples show that the edge adaptive wavelet transform achieves good reproduction of sharp edges even at very low bit rate. Because of the additional information, the PSNR of the compressed images is typically slightly lower than that of standard coders, but the subjective vision impression is improved significantly.2681|198|0更新时间：2024-05-07
-
- 摘要：The compression of point cloud is one step of the necessary preprocessing in the reverse engineering modeling. The issue of detection and preservation of boundary points, however, has not been usually considered in many point cloud compression algorithms. Then, boundary points could be withdrawn during large scale compression, and integrality of data could not be guaranteed. In this paper, we provide an algorithm to detect the boundary points. Obviously, if a point is a boundary point, then its circumambient points will distribute only on one side or around a corner. If a point is not a boundary point, its neighboring points will distribute symmetrically. By this way, boundary points will be detected by analyzing the relation of points in a trivial neighboring region. The algorithm will be effective for inner boundary points as well as outers. It sorts the boundary points further, and detects the transitional points in boundary points, at last constructs boundary polygon lines by transitional points. The distance between a boundary point and its neighboring boundary points is used to detect the transitional points. Therefore, the algorithm not only can satisfy preserving the boundary points in a large scale compression, but also prepare for reversion modeling by boundary polygon lines.3863|196|0更新时间：2024-05-07
-
-
- 摘要：In the field of computer vision, the research on human detection has a wide application prospect. Prevalent human detection methods usually use traditional statistical theory, which is based on empirical risk minimization. But the minimization of empirical risk over limited training data does not imply good generalization to novel test data. Aiming at the shortcomings of traditional statistical theory used in human detection, a new method based on SVM is presented in this paper. An adaptive background subtraction method combined with color is used to segment the motion objects. Then the trained SVM classifier distinguishes the motion object whether it is a human or not. In order to simplify the design of SVM classifier and improve efficiency of machine learning, a center radiating vector representation is proposed to abstract features of the object. And the optimal representation is obtained by experiments. During the machine learning, a bootstrap method is adopted to reduce the complexity of training SVM. Experiments show that the performance of SVM is better than ANN, and the radial basis function SVM has better performance than other SVMs on human distinguish. This method has strong robustness and high accuracy.2568|173|0更新时间：2024-05-07
- 摘要：In the video detection system of urban traffic flow, it is a hard work to detect vehicle at night. The traditional methods are based on vehicle headlights and headlight beams using mathematical morphology, they are sensitive to environment illumination with high computing cost. This paper presents a method based on color and motion information. Firstly, the algorithm detects and tracks vehicle taillights using color information on the images, then matches the detected vehicle taillights using motion information and priori knowledge, finally vehicle flow is computed. As shown in experiments, the algorithm can exactly detect moving vehicles at night and it is robust to rainy conditions.3844|245|0更新时间：2024-05-07
- 摘要：This paper proposed a real time vehicle speed measurement algorithm based on video image monitoring. According to the speed equation, to get the speed, if we have two images which the acquisition time between them is known, we need to look for a way to get the position of same block in the two images. In this paper, the theory, implementation flowchart and the key points of the algorithm were presented. A fast vehicle lamp locating method based on vehicle lamp characteristic and horizontal gray level difference projection was provided to solve the main difficulty of the algorithm. The experimental result and the problems that could be resolved in the future were discussed in the end, and the block can be located in 13ms meantime on a conventional 800MHz Intel Pentium III. According the data of result in experiment, the accuracy of speed measurement could probably achieve more than 92%. It shows the possibility to offer a real time system to measure speed of vehicle by video only. Because of the simple algorithm, the system can be implemented on a wide range of small low power devices, including embedded processors.4403|343|0更新时间：2024-05-07
- 关键词：Snake model;active contour model;video object tracking;histogram2907|193|0更新时间：2024-05-07
- 摘要：Hand petal of intravenous needles for single use is assembled by hand in most enterprises in China, but there are many shortcomings. In this paper, an image description method of identifying intravenous needles images by Hu moment invariants is studied. Images of intravenous needles of different angles are obtained by means of backlighting, which is able to acquire the intact and available information of outline of intravenous needles. The experiment proves that with the increase of angle, the moment invariants of the images are changed by the rules. Hu moment invariants are feasible and effective to describe the images of intravenous needles for single use.2358|157|0更新时间：2024-05-07
- 摘要：Window technique is a universal method of image display in CT radiology clinics. Due to the CT number of various tissues, organs or structures being different in a variety of dynamic ranges, the operations of reasonably selecting and manually adjusting Window Level/Width are required in order to observe the details of structures in an optimal display. Otherwise, the fiber of an organism or the focus can not be displayed clearly, even resulting in the diagnostic information lost, films wasted and mistakes in diagnosis. The proper operations depend on the clinical experiences. And the manual performing causes lots of inconvenience for radiologists to analyze images and diagnose. In the paper, a new method of CT medical images contrast enhancement is represented based on the multiscale edge representation of signals proposed by Mallat and Zhong. Comparing with the traditional Window technique, our method is adaptive, and the manual operation to adjust properly Window Level / Width can be avoided. In addition, this method can enhance the contrast of organisms whose intensity are different less, stress edge contours of an organism much more, and display fine and tiny structures more clear. These properties make our method meet the demand of radiology clinics much better, and indicate the practicality and potential application value of the proposed method in the field of CT medical images contrast enhancement.2568|189|0更新时间：2024-05-07
-
- 摘要：The high dimensions of hyperspectral remote sensing image have caused problems to further processing. In order to solve the above problems, this paper proposes an adaptive band selection (ABS) algorithm of dimensional reduction, which selects high informative and low correlative bands by calculating the index of each band. After calculating and rearranging the index of each band, there are two methods of selecting the final bands: one is to select the bands whose index is bigger than the specified index, another is to select the first n bands whose index is the first n bands. In order to testify the effect of ABS method, Bayesian supervised classification was implemented on the dimensionally reduced image. The results of the classification show that highly informative bands can be selected by ABS method, the classification accuracy of selected bands is 10 4% bigger than the original image, and the computing complication is decreased rapidly too.6236|213|0更新时间：2024-05-07
- 摘要：Based on a data dimension compression technique, this paper studies the principles and method of remote sensing information extraction for urban built up land. With the detailed analysis of urban land use types, the study selects three indices,i.e. Normalized Difference Built up Index(NDBI), Modified Normalized Difference Water Index (MNDWI) and Soil Adjusted Vegetation Index(SAVI), to represent three major urban land use/cover classes, including built up land, water body, and vegetation. This reduced 7 multispectral bands of a Landsat 7 ETM+ subscene of Fuzhou city to three index bands generated from the original multispectral bands and thus dramatically decreased band correlation, data redundancy and spectral confusion between different land use/cover classes. The three index bands are then used to compose a new image. A maximum likelihood based supervised classification was carried out on the new three band image and the built up land is finally extracted by masking out non built up land classes. The extraction result achieves a 91.2% overall accuracy. Therefore, the method is an effective one for the remote sensing information extraction of urban land use.3366|201|0更新时间：2024-05-07
-
- 摘要：This paper, by using the strong resistance of Hough transform to noise and its characteristics of extracting points in the vicinity of different lines, studies the curve detection when x i,y i of the discrete data point set M satisfy the exponential function relation. Firstly, the new data points M *( x * i,y * i ) are obtained by logarithmically transforming the discrete data points, and the relation between x * i and y * i becomes linear. Secondly, Hough transform is used to detect the line of M * so that the parameters of lines are obtained. Thirdly, the parameters of lines obtained by Hough transform are used to calculate the distances d ki from the points to the line, and compared d ki with the given threshold d k , so the points in the vicinity of the different lines are extracted and the interferences or noise of data point set M * are deleted. At last, by using the fit line of the least square method and through inverse transform, parameters a and b of the fit curve equation after interferences or noise being deleted are obtained. This paper proposes a new method of detecting the exponential functional curves, which overcomes the three problems existed in the use of the fit curve of the least square method and does not need high accuracy of Hough transform.4739|179|0更新时间：2024-05-07
-
- 摘要：Triangulations from 3D scattered point sets play an important role in surface modeling. The applied criterion determines the final triangulation. Since curvature is one of the intrinsic properties of a surface, an algorithm for estimating the curvature of a convex quadrilateral is given in this paper. Then a new criterion of minimizing this curvature is proposed. Comparisons between this minimal curvature criterion and other criteria are concisely enunciated through detailed analyses to an example. The results indicate that the triangulations based on the criterion have the tendency to hold better geometric properties. This has practical value for surface reconstructions and surface designs etc.3102|182|0更新时间：2024-05-07
- 摘要：Shape recognition of planar curves is a fundamental problem of pattern recognition. But it has not been well resolved by now. It is difficult to value the shape difference between two curves. This paper gives a mathematical definition to that two open planar curves have the same shape based on curvature representation, therefore, a new kind of shape distance is proposed. It is concluded that the problem of computing the shape distance between two open planar curves can be changed to the problem of a functional extremum, and the differential equation is given, which can be used to figure out the shape distance. However, it is very difficult to solve that equation. Since there is no better means at hand, a rude method of fragmentation matching is used in experiments. It is also presented in this paper how to implement the algorithm step by step, especially how to represent an open planar curve in the discrete case. A constant momentum based Gaussian filtering is applied to reduce the noise sensitivity of curvature based representation. The experimental results show that this kind of shape distance is effective.3160|154|0更新时间：2024-05-07
- 摘要：Audio stream needs to keep continuous, and video stream movement should be consistent with audio stream in Video Conference System over IP. Varieties of net conditions bring transmission delay jitter. The receiver must take synchronization control to resume time relation of stream before playing them according to the facts. This paper discusses the synchronization control in convergence MCU mode and present some specific and operatable solutions.3393|191|0更新时间：2024-05-07
- 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