This paper presents a new kind of image retrieval system which obtains the feature vectors of images by estimating their fractal dimension; and at the same time establishes a tree structure image database. After prep...This paper presents a new kind of image retrieval system which obtains the feature vectors of images by estimating their fractal dimension; and at the same time establishes a tree structure image database. After preprocessing and feature extracting, a given image is matched with the standard images in the image database using a hierarchical method of image indexing.展开更多
New adaptive preprocessing algorithms based on the polar coordinate system were put forward to get high-precision corneal topography calculation results. Adaptive locating algorithms of concentric circle center were c...New adaptive preprocessing algorithms based on the polar coordinate system were put forward to get high-precision corneal topography calculation results. Adaptive locating algorithms of concentric circle center were created to accurately capture the circle center of original Placido-based image, expand the image into matrix centered around the circle center, and convert the matrix into the polar coordinate system with the circle center as pole. Adaptive image smoothing treatment was followed and the characteristics of useful circles were extracted via horizontal edge detection, based on useful circles presenting approximate horizontal lines while noise signals presenting vertical lines or different angles. Effective combination of different operators of morphology were designed to remedy data loss caused by noise disturbances, get complete image about circle edge detection to satisfy the requests of precise calculation on follow-up parameters. The experimental data show that the algorithms meet the requirements of practical detection with characteristics of less data loss, higher data accuracy and easier availability.展开更多
A 320×240 CMOS image sensor is demonstrated,which is implemented by a standard 0.6 μm 2P2M CMOS process.For reducing the chip area,each 2×2-pixel block shares a sample/hold circuit,analog-to-digital convert...A 320×240 CMOS image sensor is demonstrated,which is implemented by a standard 0.6 μm 2P2M CMOS process.For reducing the chip area,each 2×2-pixel block shares a sample/hold circuit,analog-to-digital converter and 1-b memory.The 2×2 pixel pitch has an area of 40 μm×40 μm and the fill factor is about 16%.While operating at a low frame rate,the sensor dissipates a very low power by power-management circuit making pixel-level comparators in an idle state.A digital correlated double sampling,which eliminates fixed pattern noise,improves SNR of the sensor, and multiple sampling operations make the sensor have a wide dynamic range.展开更多
Underwater terrain-aided navigation is used to complement the traditional inertial navigation employed by autonomous underwater vehicles during lengthy missions. It can provide fixed estimations by matching real-time ...Underwater terrain-aided navigation is used to complement the traditional inertial navigation employed by autonomous underwater vehicles during lengthy missions. It can provide fixed estimations by matching real-time depth data with a digital terrain map, This study presents the concept of using image processing techniques in the underwater terrain matching process. A traditional gray-scale histogram of an image is enriched by incorporation with spatial information in pixels. Edge comer pixels are then defined and used to construct an edge comer histogram, which employs as a template to scan the digital terrain map and estimate the fixes of the vehicle by searching the correlation peak. Simulations are performed to investigate the robustness of the proposed method, particularly in relation to its sensitivity to background noise, the scale of real-time images, and the travel direction of the vehicle. At an image resolution of 1 m2/pixel, the accuracy of localization is more than 10 meters.展开更多
The Image sensor needs various image processing by Image Signal Processor （ISP） to improve image quality. Conventional cameras have their own software ISP functions to perform in PC instead of using commercial ISP c...The Image sensor needs various image processing by Image Signal Processor （ISP） to improve image quality. Conventional cameras have their own software ISP functions to perform in PC instead of using commercial ISP chips. However these methods have problems such as large computation for image processing. In this paper, th authors proposed ISP that significantly reduced chip area by efficiently sharing of hardware and software. Large operation blocks are designed to hardware for high performances, and hardware is imployed simultaneously with software considering the size of the hardware. The implemented ISP can process Video Graphics Array （VGA） （640 ＊ 480） images and has 91 450 gates size in 0. 35 μm process.展开更多
A novel algorithm for skeleton extraction is proposed in the paper. By numbering objeet＇s border dements on spatial position, the border gap （BG） of inner pixel of the object is calculated; an 8-connected medial-ax...A novel algorithm for skeleton extraction is proposed in the paper. By numbering objeet＇s border dements on spatial position, the border gap （BG） of inner pixel of the object is calculated; an 8-connected medial-axis hierarchy is derived by the BG; a thinning method including slicing and counting is proposed to improve the processing speed; branches with minor importance are truncated by vector diversity Vd and length-width ratio （LWR） with support vector machine （SVM） classifier. Experiments demonstrate that the derived skeletons keep good connectivity, especially in long and narrow area.展开更多
Mutual information （MI）-based image registration is effective in registering medical images, but it is computationally expensive. This paper accelerates MI-based image registration by dividing computation of mutual ...Mutual information （MI）-based image registration is effective in registering medical images, but it is computationally expensive. This paper accelerates MI-based image registration by dividing computation of mutual information into spatial transformation and histogram-based calculation, and performing 3D spatial transformation and trilinear interpolation on graphic processing unit （GPU）. The 3D floating image is downloaded to GPU as flat 3D texture, and then fetched and interpolated for each new voxel location in fragment shader. The transformed resuits are rendered to textures by using frame buffer object （FBO） extension, and then read to the main memory used for the remaining computation on CPU. Experimental results show that GPU-accelerated method can achieve speedup about an order of magnitude with better registration result compared with the software implementation on a single-core CPU.展开更多
This paper briefly introduces the characteristics and structure of symbol QR two-dimensional code, a detailed analysis of the image processing method to identify QR code of the whole process, and the bilinear mapping ...This paper briefly introduces the characteristics and structure of symbol QR two-dimensional code, a detailed analysis of the image processing method to identify QR code of the whole process, and the bilinear mapping method is applied to image correction, the final steps of decoding are given. The actual test results show that, the design algorithm has theoretical and practical, this recognition system can correctly read QR code, and has high recognition rate and recognition speed, has practical value and application prospect.展开更多
This study presents a time series prediction model with output self feedback which is implemented based on online sequential extreme learning machine. The output variables derived from multilayer perception can feedba...This study presents a time series prediction model with output self feedback which is implemented based on online sequential extreme learning machine. The output variables derived from multilayer perception can feedback to the network input layer to create a temporal relation between the current node inputs and the lagged node outputs while overcoming the limitation of memory which is a vital port for any time-series prediction application. The model can overcome the static prediction problem with most time series prediction models and can effectively cope with the dynamic properties of time series data. A linear and a nonlinear forecasting algorithms based on online extreme learning machine are proposed to implement the output feedback forecasting model. They are both recursive estimator and have two distinct phases： Predict and Update. The proposed model was tested against different kinds of time series data and the results indicate that the model outperforms the original static model without feedback.展开更多
文摘This paper presents a new kind of image retrieval system which obtains the feature vectors of images by estimating their fractal dimension; and at the same time establishes a tree structure image database. After preprocessing and feature extracting, a given image is matched with the standard images in the image database using a hierarchical method of image indexing.
基金Project(20120321028-01)supported by Scientific and Technological Key Project of Shanxi Province,ChinaProject(20113101)supported by Postgraduate Innovative Key Project of Shanxi Province,China
文摘New adaptive preprocessing algorithms based on the polar coordinate system were put forward to get high-precision corneal topography calculation results. Adaptive locating algorithms of concentric circle center were created to accurately capture the circle center of original Placido-based image, expand the image into matrix centered around the circle center, and convert the matrix into the polar coordinate system with the circle center as pole. Adaptive image smoothing treatment was followed and the characteristics of useful circles were extracted via horizontal edge detection, based on useful circles presenting approximate horizontal lines while noise signals presenting vertical lines or different angles. Effective combination of different operators of morphology were designed to remedy data loss caused by noise disturbances, get complete image about circle edge detection to satisfy the requests of precise calculation on follow-up parameters. The experimental data show that the algorithms meet the requirements of practical detection with characteristics of less data loss, higher data accuracy and easier availability.
文摘A 320×240 CMOS image sensor is demonstrated,which is implemented by a standard 0.6 μm 2P2M CMOS process.For reducing the chip area,each 2×2-pixel block shares a sample/hold circuit,analog-to-digital converter and 1-b memory.The 2×2 pixel pitch has an area of 40 μm×40 μm and the fill factor is about 16%.While operating at a low frame rate,the sensor dissipates a very low power by power-management circuit making pixel-level comparators in an idle state.A digital correlated double sampling,which eliminates fixed pattern noise,improves SNR of the sensor, and multiple sampling operations make the sensor have a wide dynamic range.
基金Supported by the National Natural Nature Science Foundation of China （Grant No. 41376102）, Fundamental Research Funds for the Central Universities （Gant No. HEUCF150514） and Chinese Scholarship Council （Grant No. 201406680029）.
文摘Underwater terrain-aided navigation is used to complement the traditional inertial navigation employed by autonomous underwater vehicles during lengthy missions. It can provide fixed estimations by matching real-time depth data with a digital terrain map, This study presents the concept of using image processing techniques in the underwater terrain matching process. A traditional gray-scale histogram of an image is enriched by incorporation with spatial information in pixels. Edge comer pixels are then defined and used to construct an edge comer histogram, which employs as a template to scan the digital terrain map and estimate the fixes of the vehicle by searching the correlation peak. Simulations are performed to investigate the robustness of the proposed method, particularly in relation to its sensitivity to background noise, the scale of real-time images, and the travel direction of the vehicle. At an image resolution of 1 m2/pixel, the accuracy of localization is more than 10 meters.
基金sponsored by ETRI System Semiconductor Industry Promotion Center,Human Resource Development Project for SoC Convergence and“System IC2010”project of Korea Ministry of Knowledge Economy
文摘The Image sensor needs various image processing by Image Signal Processor （ISP） to improve image quality. Conventional cameras have their own software ISP functions to perform in PC instead of using commercial ISP chips. However these methods have problems such as large computation for image processing. In this paper, th authors proposed ISP that significantly reduced chip area by efficiently sharing of hardware and software. Large operation blocks are designed to hardware for high performances, and hardware is imployed simultaneously with software considering the size of the hardware. The implemented ISP can process Video Graphics Array （VGA） （640 ＊ 480） images and has 91 450 gates size in 0. 35 μm process.
文摘A novel algorithm for skeleton extraction is proposed in the paper. By numbering objeet＇s border dements on spatial position, the border gap （BG） of inner pixel of the object is calculated; an 8-connected medial-axis hierarchy is derived by the BG; a thinning method including slicing and counting is proposed to improve the processing speed; branches with minor importance are truncated by vector diversity Vd and length-width ratio （LWR） with support vector machine （SVM） classifier. Experiments demonstrate that the derived skeletons keep good connectivity, especially in long and narrow area.
基金Supported by National High Technology Research and Development Program("863"Program)of China(No.863-306-ZD13-03-06)
文摘Mutual information （MI）-based image registration is effective in registering medical images, but it is computationally expensive. This paper accelerates MI-based image registration by dividing computation of mutual information into spatial transformation and histogram-based calculation, and performing 3D spatial transformation and trilinear interpolation on graphic processing unit （GPU）. The 3D floating image is downloaded to GPU as flat 3D texture, and then fetched and interpolated for each new voxel location in fragment shader. The transformed resuits are rendered to textures by using frame buffer object （FBO） extension, and then read to the main memory used for the remaining computation on CPU. Experimental results show that GPU-accelerated method can achieve speedup about an order of magnitude with better registration result compared with the software implementation on a single-core CPU.
文摘This paper briefly introduces the characteristics and structure of symbol QR two-dimensional code, a detailed analysis of the image processing method to identify QR code of the whole process, and the bilinear mapping method is applied to image correction, the final steps of decoding are given. The actual test results show that, the design algorithm has theoretical and practical, this recognition system can correctly read QR code, and has high recognition rate and recognition speed, has practical value and application prospect.
基金Foundation item： the National Natural Science Foundation of China （No. 61203337）
文摘This study presents a time series prediction model with output self feedback which is implemented based on online sequential extreme learning machine. The output variables derived from multilayer perception can feedback to the network input layer to create a temporal relation between the current node inputs and the lagged node outputs while overcoming the limitation of memory which is a vital port for any time-series prediction application. The model can overcome the static prediction problem with most time series prediction models and can effectively cope with the dynamic properties of time series data. A linear and a nonlinear forecasting algorithms based on online extreme learning machine are proposed to implement the output feedback forecasting model. They are both recursive estimator and have two distinct phases： Predict and Update. The proposed model was tested against different kinds of time series data and the results indicate that the model outperforms the original static model without feedback.
