最新刊期
卷 19 ， 期 1 ， 2014
- 摘要：For a segmentation method to be useful it must be fast, easy to use, and produce high quality segmentations. It is clear that only few algorithms can offer this combination under various conditions. Thresholding is a popular image segmentation method that converts a gray level image into a binary image, and it has been widely applied in many fields. However, uncertainty is an inherent part of image thresholding in real world applications, and the automatic selection of optimum thresholds is still a challenge. In order to select the optimal threshold for image segmentation, an adaptive method based on rough set is proposed. The proposed method analyzes the rough set-based framework for image representation, establishes the relation between image rough granularity and local grayscale standard deviation, and obtains the optimal partition granularity by minimizing the adaptive rough granularity criteria. Next, the method defines the upper and lower approximation for image object and background, as well as the corresponding rough measure, and then produces the optimal grayscale by searching gray levels to maximize rough entropy. Finally, taking the boundary of object and background as transition region, the method achieves image thresholding according to the mean grayscale of the transition region. We have developed a program for the proposed method using MATLAB. The proposed method needs no input parameter, and its time complexity is approximately linear related to the size of the original image. Theoretically, the proposed method is efficient, and the segmented result is produced within 5 seconds for an image with the length and width of 256 in practice. By three groups of experiments on a variety of synthetic and real images, whose gray level are all 256, the performance of the proposed method is compared with the published results from three traditional state-of-art algorithms, and a rough set-based algorithm is also involved in the comparison, which have also been implemented under MATLAB 2007b environment. On one hand, we provide a qualitative comparison of our output against these relative algorithms. Compared with the rough set-based algorithm, the experimental results suggest the proposed method is effective to yield the approximately ideal results, that is, ground-truth images. On the other hand, the quantitative results are also reported using five measure metrics, including misclassification error, mean structural similarity, false negative rate, and false positive rate, which always lies between 0 and 1. Compared with three traditional state-of-art algorithms, the proposed method outperforms the other methods, demonstrating the highest score in most cases. The proposed method obtains the adaptive window size to construct rough set related with the upper and lower approximation for the image object and the background, and then segments the image based on the transition region determined by the rough set. These processes are proved to improve the segmentation result. In summary, it is indicated by the quantitative and qualitative experiments that, the proposed method performs good and robust image thresholding results, especially for non-destructive testing images. We can conclude that, our technique based on rough entropy has preferable adaptive performance and is superior to other existing methods. The proposed method is reasonable and effective, and can be as a powerful alternative to the traditional methods.关键词：image segmentation;rough set;rough granularity;rough entropy;transition region-based thresholding5282|574|0更新时间：2024-05-07
- 摘要：A key question of multi-focus image fusion is how to decide the clarity of the source images. To solve this problem, a new focus measure is proposed. The new focus measure based on properties of local extrema is proposed, namely, the normalized structure extrema. We propose a fast image fusion method based on the new focus measure and a quick estimation of the fusion decision matrix. With the new criterion of image clarity and the fusion method introduced in the paper, multi-focus image fusion yield superior performance than the conventional approaches in terms of the test results. Normalized structure extrema number is a new focus measure that can judge the clarity of a region effectively and more robustly against pulse noise. Compared with some traditional methods of multi-focus image fusion on both the subjective visual effect and objective evaluation, it indicates that the presented method improves the fused results and efficiency saliently.关键词：image fusion;multi-focus image;extrema;clarity of image4166|977|0更新时间：2024-05-07
- 摘要：Aiming at the camera pose estimation problem, with the corresponding points number greater than six, we propose an accurate and fast algorithm which has a time complexity of () and is applicable to both: calibrated cameras and uncalibrated cameras. First, four non-coplanar virtual control points are selected and linear equations according to the relationship between space points and virtual control points and the image of space points are being setup. Then the image coordinates of the virtual control points and the camera internal parameters are solved by these equations. Second, the rotation matrix and the translation vector are solved by the POSIT (pose from orthography and scaling with iterations) algorithm according to virtual control points and their image points. Experimental results with simulative data and real images indicate that the time complexity and calculation precision of the proposed algorithm are better than EPnP (efficient perspective-n-point) algorithm which is the existing accurate and fast pose estimation algorithm for calibrated camera.The proposed algorithm can estimate both camera internal and external parameters, and time complexity and calculation precision are better then existing methods.关键词：calibrated camera;uncalibrated camera;PnP problem;EPnP algorithm;POSIT algorithm6556|399|0更新时间：2024-05-07
- 摘要：According to the problem of embedding capacity and visual quality, a lossless information hiding method based on reversible integer transform algorithm with larger embedding capacity for image is proposed. By the proposed reversible integer transform algorithm, the average value of an image block with n pixels is calculated first, and the difference between every pixel's value with the mean value is expended 4 times as large as before during the integer transform. The two least significant bits of all the transformed pixels' values are equal, producing redundancy information which can be used to embed 2(-1)bits data. Based on the proposed new reversible integer transform algorithm, an information hiding method with larger embedding capacity for image was presented, the experimental results showed that the proposed method has lager embedding capacity and better security, and both the original host image and the embedded data could be restored without any distortion from the marked image.关键词：reversible integer transform;lossless information hiding;difference expansion;maximal payload4205|310|0更新时间：2024-05-07
- 摘要：Under the framework of the Bayesian inference,tracking methods based on sparse representations can deal with complex appearance changes in the video scene successfully and robustly.However,the computation costs are too expensive to achieve real-time tracking.To solve this problem,a new real-time tracking method based on L-norm minimization is proposed in this paper.The proposed method introduces the L norm minimization into the PCA reconstruction,removes trivial templates from the sparse tracking method and presents an effective object representation model based on the L-norm minimization.An observation likelihood function that takes occlusion into account is designed in this paper.The experiments on many challenging image sequences demonstrate that the proposed method achieves the same and even better results when compared with several state-of-the-art tracking algorithms. Furthermore, it runs fast with a speed of about 20 frames/s.The proposed method in this paper can handle occlusion,illumination changes,scale changes and no-rigid appearance changes effectively in video surveillance scenes with a lower computation complexity.Additionally, it can run in real-time.关键词：object tracking;sparse representation;Bayesian inference;L;-norm minimization3941|457|0更新时间：2024-05-07
- 摘要：Intelligent transportation is a service system based on the modern electronic information technology for transportation. With the development of intelligent traffic system, video surveillance technology is often used in this area. While using the video surveillance system to detect the traffic scene, the false detections often occur when two or more vehicles approaches each other. This phenomenon increases the difficulty of vehicle detection. To overcome this problem, it is necessary to establish a reliable, practical splitting mechanism for the occluded vehicles. Further more, we proposed a new identification and segmentation algorithm for occluded vehicles. In order to obtain the moving areas, we use the background differencing to detect the current image. Meanwhile, t shadow areas need to eliminated. We use a shadow detection algorithm that is based on the HSV space features. Then, the image is divided into blocks to reduce the processing time. We use the width/height ratio and occupancy ratio to judge whether a moving area contains one or more vehicles. To find out the recessed area between vehicles, a new algorithm called the "seven grids" is presented. The new "seven grids" algorithm is a matrix of seven rows and seven columns. First, the recessed area detection algorithm calculates the edge of the vehicle regions. Then it uses the "seven grids" to traverse all the edges, puts the detected point at the center of the "seven grids" and determines for each point whether it is a concave point. At last, it determines the recessed areas based on the concave points, and finds the occluded areas between vehicles by matching the corresponding recessed areas. At the same time, there are some differences about the color and brightness between different vehicles. This situation is more obvious when the occlusion phenomenon exists between vehicles. Therefore, we use the algorithm for edge detection to detect these occluded areas. To identify edges of vehicles which are regarded as segmentation curves. Then we use them to segment the occluded vehicles. Traditional segmentation methods often work to find the segmentation points and connect the corresponding segmentation points to segment occluded vehicles. This method can effectively segment them, but the segmentation results are not accurate. The segmentation method we propose is committed to find the occluded area, and uses the edges of vehicles into the occluded area to segment occluded vehicles. The algorithm satisfies the real-time requirement and can effectively segment the occluded vehicles. Compared with other methods, it has better segmentation results and higher recognition success rate, and the recall and precision can reach up to 90%. In this paper we focus on the effectiv identification and segmentation of occluded vehicles. We propose a segmentation method based on the "seven grids". The method segments occluded vehicles effectively, and it has strong adaptability, because it does not need any prior knowledge. Experimental results demonstrate that, this occluded vehicles detection algorithm has a high recognition rate. The proposed vehicles segmentation method can segment the overlapped ones accurately and completely.关键词：concavity analysis;seven grids;recessed area;occluded area;vehicle splitting3806|637|0更新时间：2024-05-07
- 摘要：Calibration of hand structures is one of important steps to create an individualized three-dimensional hand model and estimate hand-pose. Because the hand structure is covered by the hand surface, the common cameras cannot directly get the hand structure, and people were required to participate in the calibration of hand structure, which cannot be completely processed by the computer. That makes the calibration complex and more time is needed. In order to acquire an automatic calibration of a hand structure, a set of image segmentation and enhancement techniques were used to find the link structure of the hand, based on the relationship of hand prints and joints in the individual's hand. The techniques of 3D modeling using multiple cameras, 3D fitting, and the Particle Swarm Optimization algorithm were used to optimize the link structure of hand. Proved by the experiments, the calibration method proposed in this paper can successfully extract the structure of the hand, and each stage of the calibration process doesn't need people participating in. Comparing with the original method, the calibration method proposed in this paper is completed by the computer. As a result the calibration time is saved and errors are reduced.关键词：individuation;link structure of hand;voxel data;hand model;hand-pose estimation3718|265|0更新时间：2024-05-07
- 摘要：Edges are one of the most important features of an image, they are the basis of many image analysis and recognition techniques. The continuity and noise immunity of the edge extraction is particularly important for the segmentation and measurement. Regional growth algorithms can be used to extract the target area. They can provide the nece-ssary support for the matting and statistical measurement. For the purpose of effective contour extraction, we propose a method of image edge extraction combined with a Gaussian weighted distance graph in this paper. First, by calculating the distance between the pixels within the sub-block regions, the graph of Gaussian weighted distances is obtained. Comparing with the original figure, it not only can better highlight the edge contour, but also can get a uniformed background gray. Second, by analyzing the histogram of the Gaussian weighted distance, the gray values can be divided into two classes, each class center is calculated for active contour without edge (CV) parameters of and . Finally, edges of the image are found using the CV model. Comparing with other edge extraction algorithms, the proposed algorithm not only has better noise immunity, but also can guarantee the continuity of the image edge extraction. The experimental results demonstrate the effectiveness of the proposed algorithm.关键词：Gaussian weighted distance;edge extracting;CV model;gray histogram;class center6505|917|0更新时间：2024-05-07
- 摘要：In the field of direct part marking technology, 2D barcode is more and more widely used, but there is a problem that 2D barcodes marked on a cylindrical part have spatial distortions which will result in reading difficulties. Existing algorithms mostly attempt to recognize 2D barcode marked on the plane which only have perspective and incline distortion. However, they can hardly recognize 2D barcode marked on cylindrical part because the pattern of 2D barcode is changed by spatial distortion. For correcting this difficulty, a new method should be proposed to recovery the distorted 2D barcode, and the radius parameter of cylindrical part should be automatically got without manual measurement.We firstly analyze the principle of forming distorted 2D barcode marked on cylindrical parts, and we establish a spatially distorted 2D barcode correction model based on calibrated machine vision mechanism and the principle of perspective projection. then deduces the relationship between fitted ellipse and the radius of cylindrical part in the image, in addition, estimates the radius by the ellipse parameters. finally substitutes it into the correction model to get the corrected 2D barcode.In the experiment, we select two distorted 2D barcodes marked on the cylindrical parts with different radius parameter, and acquires pictures of them from different views. It gets fixed parameters of the machine vision device, based on that the spatially distorted 2D barcode correction model could be established.The experiment results show that, the two distorted 2D barcodes are recovered well and the processing time is limited in 1 second.Compared with previous methods, this algorithm is more concise because of prior calibration and spatial distortion mode. Besides, this algorithm is more automated because the radius parameter could be automatically got by the image.Based on camera calibration, the method is effective for correcting spatially distorted 2D barcode marked on arbitrary radius cylinder.The method extends the application range of 2D barcode that not only flat part but also cylindrical part could also be identified. However, in the future work it should introduce nonlinear distortion of the lens and get the more accurate camera calibration parameters for more ideal restoration result.关键词：cylindrical part;spatially distorted 2D barcode;spatial distortion;parameter calibration;spatial distortion correction4088|267|0更新时间：2024-05-07
- 摘要：In inertia confinement fusion (ICF) experiments, the visual measurement accuracy and speed of targets directly affect the success rate. Since important imaging feature of several kinds of targets are elliptical, a fast and effective ellipse detection algorithm is required for the targets visual measurement. However, the vague contours and irregular brightness of target images pose great challenges to conventional ellipse detection approaches. Meanwhile, considering the real-time requirement, a new fast ellipse detection algorithm is proposed to solve the problems mentioned above. The ellipse detection process includes two parts: extracting edge pixels of ellipses, and then extracting the ellipse parameters fitting these edge pixels. In order to obtain accurate edges pixels of ellipse, a feature that edge pixels of ellipse have large gray change rate in the direction, which connects the edge pixel and the center of ellipse, is used. The new ellipse-edge detection algorithm is called polar coordinates edge detection (PCED). First, a downscaling target image is used to pre-estimate the center of the ellipse with a conventional ellipse detection method. Second, a polar coordinates system is built, and the origin point of the system is the pre-estimate center of the ellipse. Last, PCED finds the pixels with extreme gray change rate in the ray, which starts from the origin point of the polar coordinates system, as the detected edge pixels. Furthermore, PCED keeps the edge detection in the local region of an image to guarantee the real-time requirement and the effectiveness. Once the edge pixels are detected by PCED, an adaptive ellipse parameters extraction algorithm based on RANSAC is adopted to get the ellipse parameters fitted to the edge pixels. The proposed ellipse parameters extraction algorithm adopts cluster analysis in ellipse parameters space to choose the optimal estimated ellipse parameters. Then, the consistent pixels of the optimal estimated ellipse parameters are chose from the detected ellipse edge pixels. Finally, the results of the ellipse parameters are calculated by the consistent pixels using least square method. The results of comparison experiment between PCED algorithm and Canny algorithm show that the PCED algorithm could achieve more accurate and more effective ellipse edge pixels compared with Canny algorithm, which also makes the following ellipse parameters extracting process more easily. The experiment results of ellipse detection for practical target images show that the processing speed of the proposed algorithm is about 110 ms for one image, which is a significantl increase compared to other conventional algorithms. Moreover, the proposed algorithm also has good performance in repeatability and consistency test experiments. First, the proposed ellipse detection algorithm uses the PCED algorithm to gain the ellipse edge pixels effectively. Then, the proposed ellipse detection algorithm adopts the adaptive ellipse parameters extraction algorithm based on RANSAC to calculate the ellipse parameters fitting to the detected edge pixels. Taking the advantages of the two processes, the proposed algorithm could gain numerically small but effective ellipse edge pixels, and then could get accurate ellipse parameters fast. The comparison experiments demonstrate that the proposed method has advantages of low time consumption, high accuracy, and good robustness, compared to other conventional methods.关键词：inertia confinement fusion;targets measurement;edge detection;ellipse detection4024|258|0更新时间：2024-05-07
- 摘要：Monogenic signal analysis has been increasingly used in face recognition. However, the monogenic orientation has not been fully utilized which as an extremely important geometric information. In this paper, a novel coding method named EPMOD (enhanced patterns of monogenic orientation difference) is proposed to extract the local orientation features. Then a new face recognition method fusing MBP (monogenic binary pattern) and EPMOD is proposed. First, MBP feature and EPMOD feature are extracted by using multi-scale monogenic filter; then, BFLD (block-based Fisher linear discrimination) is used to reduce the dimensionality of the two descriptors. Finally, the two kind of feature is fused at score level. The experimental results on the ORL and CAS-PEAL face databases validate that the proposed algorithm has better performance than or comparable performance than LGBP and MBP but with lower time and space complexity. An effective facial feature extraction method is proposed in this paper, and the experimental results also show that our fusion approach can improve the recognition rate significantly.关键词：face recognition;monogenic filter;monogenic binary pattern(MBP);enhanced pattern of monogenic orientation difference(EPMOD);block-based Fisher linear discriminant(BFLD)4343|501|0更新时间：2024-05-07
- 摘要：Occlusion is an issue in the tracking field, and the handling of occlusion is a key measure for robust tracking algorithms. In this article, we propose a new algorithm of handling occlusion fragments-based tracking. When the target is partially occluded or the pose changes, the strength of the remaining effective patch of information remains credible. Our algorithm uses the Bhattacharyya coefficient of similarity measure as a candidate target piece with the corresponding template piece to track a target effectively. Using H back projection method to distinguish between occlusion and pose change. According to the different types of occlusion, it makes the appropriate handle to realize robust tracking of the target. The present experiment proposed the concept of relevance occlusion and non-relevance occlusion, increasing the reliability of algorithm for tracking. Through fragments-based tracking,and taking into account the correlation between the target and obstructions,it will make improved about the tracking efficiency significantly.关键词：fragmentation;relevance occlusion;target tracking;bhattacharyya coefficient4086|444|0更新时间：2024-05-07
- 摘要：For the defect of optical estimation, noise and the limitations of existing motion attention model lead to the computation results of motion attention, which cannot accurately reflect the conspicuous characteristics of motion and constrain further application of motion conspicuous map. To improve the computation accuracy of motion attention, we suggest a target detection algorithm based on multi-scale motion attention analysis in this paper. According to the mechanism of visual attention, spatial-temporal motion attention model is built. Then noise influence is reduced by the time filtering. In view of the visual observation of scale dependence, the video frames are decomposed in multiple scales and the motion attention is also computed in spatial multiple scales in space. On the basis of the correlation coefficient of macro block pixel, low scale, middle scale, and the original scale, the motion attention computation results are fused to obtain the final motion attention map. The test result using different videos show the algorithm is more correct for motion attention than other algorithms and it greatly improves the accuracy of the motion attention map. In order to improve the inaccurate computation of motion attention, we propose a motion attention computation algorithm based on spatio-temporal multi-scale analysis. For different complex motion video scenes, the proposed algorithm can obviously enhance the computation accuracy of motion attention and lay a good foundation for the further application of visual motion attention.关键词：motion attention;multi-scale;fusion;correlation coefficient3173|312|0更新时间：2024-05-07
- 摘要：Due to the problem that traditional latent semantic analysis (LSA) method is unable to obtain spatial distribution information of objects and discriminative information of latent topic. We propose a scene classification approach based on multi-scale spatial discriminative probabilistic latent semantic analysis (PLSA). First, it decomposes images in multiple scales using a spatial pyramid approach to obtain spatial distribution information for images. Then, the PLSA model is used to extract the latent semantic information of each local block. Next, the latent semantic features of all local blocks are concatenated with different weights to produce the multi-scale spatial latent semantic information of image. Finally, we exploit weight learning method to learn the discriminative information between different image topics and get multi-scale spatial discriminative latent semantic information of image. Afterwards, the weight information is integrated into the support vector machine (SVM) classifier to perform image classification. Experimental results on the common three scene image datasets, viz. Scene-13, Scene-15 and Caltech-101, demonstrate that our method performs much better than the existing state-of-the-art approaches. Which demonstrate the importance of spatial information and discriminative information in image classification and further verify the effectiveness and robustness of our approach.关键词：probabilistic latent semantic analysis;spatial pyramid;discriminative information;scene classification3547|266|0更新时间：2024-05-07
- 摘要：To solve the jitter and tear problems caused by single-precision floating-point errors in global scene rendering, a complete set of solutions is presented. First, a new coordinate origins is found by using the global quad-tree structure, avoiding the disadvantages of switching coordinate origins frequently; Second, the jitter problem is solved by transforming the coordinate and matrix in the new coordinate system; finally, the depth resolution is improved by using logarithmic calculation on the GPU, and the Z-Fighting problem is solved. The experiment results show that the proposed method can be a good solution to the problem caused by single-precision floating-point errors. It has strong adaptability, low complexity, and it is controllable with accuracy and efficiency advantages.关键词：global scene;GPU rendering;dynamic coordinate system;global quad-tree;depth calculation3173|289|0更新时间：2024-05-07
- 摘要：Speckle noise is a granular structure, and it occurs when a coherent source and a non-coherent detector are used to interrogate a medium. Speckle noise is an undesirable part in the ultrasound image, since it can mask the small difference in grey level and degrade the image quality. The task of despeckling is an important step for analysis and processing of ultrasound images, which is essential for automatic diagnostic techniques. The wide spread of mobile and portable ultrasound scanning instruments also necessitates that a clearer image must be obtained to the medical practitioner. A novel despeckling algorithm for medical ultrasound images is proposed, which is based on the wavelet transformation and a bilateral filter. According to the statistical properties of medical ultrasound images in wavelet domain, an improved wavelet threshold function is proposed on the basis of the universal wavelet threshold function. The proposed wavelet threshold function is obtained by multiplying the universal wavelet threshold function with an adjustable parameter. The noise-free signal and speckle noise in the wavelet domain are modeled as generalized Laplace distribution and Gaussian distribution respectively. The Bayesian maximum a posteriori estimation is applied to get a new wavelet shrinkage algorithm. The speckle noise in the high-pass component in wavelet domain of ultrasound images is suppressed by the new wavelet shrinkage algorithm. The speckle noise in the low-pass approximation component is filtered by the bilateral filter, since the low-pass approximation component of ultrasound images also contains some speckle noise. The filtered image is then obtained via inverse wavelet transform. The comparative experiments with seven other despeckling methods are conducted. Several image quality metrics are used to compare the performance of speckle reduction, such as the peak signal to noise ratio (PSNR), the structure similarity (SSIM) and Pratt's figure of merit (FoM), as well as the computational time of different methods is presented. The filtered images of the proposed algorithm get the best result by compared to the PSNR and FoM values with other seven despeckling methods. The best result of the PSNR and FoM value means that the proposed algorithm can suppress more speckle noise and the filtered image has the similar edge to witch of the noise-reference synthetic image. In the comparison of SSIM values, the proposed algorithm also gets good performance, which means that the proposed algorithm can retain a structure similar structure to the noise-free reference synthetic ultrasound image. Observing the computational time, the proposed algorithm does not have superiority in the aspect of time consuming. The experiment of clinical ultrasound breast images with lesions is also conducted, and we can find that the proposed algorithm gets a pretty good despeckling performance. Since speckle noise limits the development of automatic diagnostic technology for ultrasound images, we propose an improved despeckling algorithm on the basis of the wavelet transform and the bilateral filter. The experiments of synthetic ultrasound images and clinical ultrasound breast images show that the proposed despeckling algorithm not only has better speckle reduction than the other seven filters, but also can preserve image details such as the edge of lesions.关键词：wavelet;bilateral filter;medical ultrasound image;generalized Laplace distribution5371|748|0更新时间：2024-05-07
- 摘要：Because of a different imaging mechanism, different modality medical images provide a variety of characteristics about image quality, space, and non-overlay complementary information. In clinical use, we need to analyze the result of multimodal medical images. In order to use medical images effectively and reasonably, a medical image fusion algorithm is proposed, combining the advantages of multi-scale and multiple directions in the Contourlet transformation. First, multi-scale and multiple directions decomposition coefficients are obtained through Contourlet transformation. Second, fusion rules are proposed by analyzing the characteristics of Contourlet transformation coefficients. An optimized image fusion rule is proposed in low frequency sub-band coefficients and high frequency sub-band coefficients. For low frequency sub-band coefficients, the weighted regional variance fusion rule is adopted in view of the image detail characteristics. The high frequency sub-band coefficients are fused by a condition-weighted rule of the main image in view of the edge detail characteristics. Finally, the final fusion image is acquired through the Contourlet inverse transformation. Different fusion rules based on Contourlet transformation and different fusion methods are analyzed. The fusion results are analyzed and compared with the measurement of human visual system and objective evaluation. Compare the new fusion method with other classical fusion algorithm to confirm the advantages of the new method. The experimental results show that the proposed algorithm is effective in retaining the original images' information and reserving the edge features successfully. A medical image weighting fusion algorithm is proposed based on Contourlet transformation. Medical images, including CT and MRI, are used for the experiments. The results show that the complementary information of medical image can be highlighted and the image definition has improved significantly.关键词：medical image;image fusion;Contourlet transformation;weighted fusion algorithm4720|671|0更新时间：2024-05-07
- 摘要：Human-machine interaction is the main way for geographic hazard remote sensing interpretation at home and abroad as it is easy to use, intuitive and efficient. But there still exist certain problems, such as excessive dependence on image color, texture, shadow and other optical elements, one-sided pursuit of interpreting keys, lacking use of DEM, insufficient applications of image comprehensive analysis, spatial analysis and 3D visualization based on GIS. Thus this paper conducts research into quantitative analysis of human-machine interaction. This paper explores three methods of one-dimensional, two-dimensional and three-dimensional remote sensing interpretation based on pre-and post-disaster DEM and high spatial resolution remote sensing images, and analyses the complementary relationships among them. Then Guanling ‘6.28’ Mega Landslide in Guizhou province of China is interpreted by comprehensively using the three methods. One-dimensional elevation curve calculation gives the initial impression of landslides interpretation, two-dimensional image comparison and analysis belongs to dynamic analysis methods, and precise three-dimensional scene interpretation is quantitative calculation. During one-dimensional elevation curve calculation the possible partition frame of collapse area, landslide area and accumulation area along the curve movement is obtained, which provides spatial reference for two-and three-dimensional interpretations. The development from two-dimensional image comparison and analysis to precise three-dimensional scene interpretation shows that high spatial resolution remote sensing image interpretation of landslides has developed from qualitative analysis highly relying on man-machine interaction mode to quantitative calculation mainly based on multi-dimensional spatial analysis models.关键词：landslide;remote sensing multi-dimension interpretation;high resolution remote sensing image;one-dimensional elevation curve calculation;two-dimensional image comparison and analysis;precise three-dimensional scene interpretation4648|370|0更新时间：2024-05-07
- 摘要：To eliminate the influence of image distortions due to the topographic relief. A novel method for the topographic correction of Polarimetric SAR (PolSAR) based on the coherence matrix has been put forward. First, the imaging geometry is reconstructed firstly based on the Range-Doppler mode. Then, the fused SAR image is obtained Cloude decomposition, which is used to match the simulated image to improve the geometric correction accuracy. Finally, the topographic correction for PolSAR is completed by the projection area normalization method and polarization azimuth shift compensation. To test the proposed method, a Radarsat-2 PolSAR image in C-band of the in Donkemadi district in China has been corrected and then applied to map the glacier. The RMSE of azimuth and range direction of register is 7.765 and 14.586 pixels and the accuracy of classification after topographic correction is more than 80%, Which proves: 1) The proposal method is able to remove geometric and radiometric distortion;2) the usage of polarization SAR in C-band in glacier mapping is feasible after the topographic correction.关键词：polarimetric synthetic aperture radar;topographic correction;snow and ice mapping;image processing3250|407|0更新时间：2024-05-07
- 摘要：The methods of change detection based on high resolution images have important applications in land use/land change, water quality change, forest resource monitoring and military target detection. In this paper, we introduce a new object-oriented change detection method for two high-resolution remote sensing images, which combines neighborhood correlation images (NCI) and the minimum-redundancy-maximum-relevance feature selection (mRMR) together. Content of main experiments:In this paper, study area is located in Shunyi district, Beijing. We choose two rapideye images as experimental data and both of them are without any cloudy covering them. Comparisons with reviewed researches:These two image data both contain abundant kinds of object types while the number of object types for traditional change detection are relative small. We design three comparative experiments to verify the effectiveness of this method: 1) compare the results of using mRMR feature selection versus without using mRMR feature selection; 2) compare the results of using NCI and mRMR feature selection together versus only using NCI; 3) compare the results of using NCI and mRMR feature selection together versus only using mRMR feature selection. The results show that the effect of change detection combining NCI and the mRMR feature selection is better than only using NCI or only using mRMR feature selection, and is more superior than neither NCI nor mRMR was introduced. Meanings:The most innovative contents about this paper is the method about how to deal with features, involving how to condense features and how to deal with high dimension features. That means we can use much more knowledge of data mining and machine learning to deal with remote sensing change detection. However,we need use more high resolution images to test this method. In theory, effectiveness of this method for change detection does not be affected by using different high resolution remote sensing data sources; However, the reality needs to be further verified by experiments.关键词：remote sensing images;high-resolution;object-oriented;change detection;neighborhood correlation images;feature selection3260|377|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