Summary form only given. A recently developed synthetic aperture radar (SAR) autofocus technique called the phase gradient autofocus (PGA) algorithm is considered. it has been developed to mitigate the problem of phas...
详细信息
Summary form only given. A recently developed synthetic aperture radar (SAR) autofocus technique called the phase gradient autofocus (PGA) algorithm is considered. it has been developed to mitigate the problem of phase error compensation, which is common to all aperture synthesis imaging systems. The phase errors manifest themselves as redundant information in the reconstructed image. This invites the use of a data-driven algorithm to estimate the phase error function and perform the restorative deconvolution. The PGA algorithm exploits this redundancy to obtain a linear minimum variance estimator of the phase error. It has been demonstrated to be robust, computationally efficient, and easily implemented in standard digital signal processing hardware.< >
A parallel implementation of algorithms to achieve viewpoint invariant target recognition from laser radar images is described. The active/passive CO/sub 2/ laser radar sensor used provides pixel-registered thermal, v...
详细信息
A parallel implementation of algorithms to achieve viewpoint invariant target recognition from laser radar images is described. The active/passive CO/sub 2/ laser radar sensor used provides pixel-registered thermal, visual, relative range and artifacts and detects edges. viewpoint invariance obtained using the translation invariant property of the Fourier transform transforms 2-D rotations and scale changes into translations. This scene representation provides 2-D viewpoint invariance and is used as input to a neural network that has been trained to recognize representative views of the targets of interest. The system is implemented on the Geometric Arithmetic Parallel Processor (GAPP), a massively parallel single-instruction-multiple-data (SIMD) computer that contains a 216*384 array of processing elements. Results obtained by processing real lasers radar images are presented.< >
The authors apply a three-dimensional (3-D) linear least-squares (LS) prediction technique to the processing of digital subtraction angiography (DSA) image sequences. The main goal of this processing is the cancellati...
详细信息
The authors apply a three-dimensional (3-D) linear least-squares (LS) prediction technique to the processing of digital subtraction angiography (DSA) image sequences. The main goal of this processing is the cancellation of motion artifacts, which is a visual structured noise that appears in current DSA images. Two important issues are addressed by this technique. The first is the misregistration of the mask and the contrast image. The second is the temporal filtering of the DSA image sequence. Instead of treating these two issues separately, as in conventional DSA, the authors combine them into a 3-D LS prediction problem. Based on this approach they developed an efficient algorithm for the solution of normal equations. To match the image sequence physical characteristics, the authors have further optimized practical parameters of this algorithm. Actual patient data are used for the evaluation of this technique. Results show a significant improvement over the existing methods.< >
The paper presents an eye tracking system specifically designed for the recording of eye movements from the visitors of a museum. Eye movements are calculated from the images of a near infra-red camera viewing the eye...
详细信息
The paper presents an eye tracking system specifically designed for the recording of eye movements from the visitors of a museum. Eye movements are calculated from the images of a near infra-red camera viewing the eye during the presentation of a sequence of slides. Thereafter, the same slides are displayed with the scanning paths superimposed. Servo systems, ''intelligent'' imageprocessingalgorithms and interactive procedures have been implemented so that the eye tracker can operate on the general public without any technical supervision.
The authors propose a vLSI circuit design method which is derived from a functional decomposition of low-level imageprocessingalgorithms. In order to design high powerful architecture, the authors have been testing ...
详细信息
The authors propose a vLSI circuit design method which is derived from a functional decomposition of low-level imageprocessingalgorithms. In order to design high powerful architecture, the authors have been testing this functional approach quite extensively, and special-purpose circuits have already been developed. Finally from a broad class of various imageprocessingalgorithms and operators, the authors have defined a computational unit called a functional element. These elements are to be associated to form a mesh-connected programmable data flow computer. This data flow computer has been defined to obtain a straight mapping of the expressed algorithms on the basis of functional programming. Thus, high computing power can be easily obtained for real-time complex imageprocessing applications. Using the equivalent data flow graphs of the algorithms, this computer can help to determine whether a given combination of precisely specified IC operators would have the requested behavior, so that a special-purpose vLSI could be easily designed. Several examples of implementation onto this computer are shown in order to point out the salient concept and advantages of this approach.
The authors present new algorithms for the segmentation of speckled images, including synthetic aperture radar (SAR) images. These algorithms are based on a hierarchical random field model proposed for speckled images...
详细信息
The authors present new algorithms for the segmentation of speckled images, including synthetic aperture radar (SAR) images. These algorithms are based on a hierarchical random field model proposed for speckled images and a maximum a posteriori (MAP) segmentation using simulated annealing. Different versions of the model are developed to represent single or multilook and intensity or complex speckled images. Parameters of the speckle processes are estimated as part of the algorithm. The segmentation algorithms are implemented on a wide class of synthetically generated and actual speckled images. The results verify that significant improvement is obtained when multilook and complex image data are used for the segmentation.
The machine vision algorithms required to increase the accuracy of a manipulator to that of its resolution are presented. This high accuracy is required for laser programming of reconfigurable integrated circuits. An ...
详细信息
The machine vision algorithms required to increase the accuracy of a manipulator to that of its resolution are presented. This high accuracy is required for laser programming of reconfigurable integrated circuits. An application-specific integrated circuit has been fabricated that performs the time-critical aspects of the image recognition algorithms at video rates.
Many problems of geophysics, imageprocessing and time series analysis involve the problem of solving Toeplitz systems of equations. We present a fast parallel O (mn) algorithm that solves both square and over -determ...
详细信息
A large number of remote sensing studies in the Netherlands are performed using airborne sensors. At the present moment, a multitude of different sensors are available, covering a wide variety of application purposes....
详细信息
The authors attempt to summarize important parameters and procedures for verifying the measurement and recognition (classification) accuracy of machine vision algorithms/systems;to alert the machine vision research co...
详细信息
The authors attempt to summarize important parameters and procedures for verifying the measurement and recognition (classification) accuracy of machine vision algorithms/systems;to alert the machine vision research community to the current, very inadequate practice in this important area;and to propose some measures to improve this situation. Two example applications from the authors' practice are given in order to illustrate experimental verification procedures presented here. The motivation for the present work is based on the fact that machine vision systems are very hard to model or simulate accurately. It is concluded that realistic large-scale experiments are the most reliable means of verifying the measurement and recognition accuracy. However, it is claimed that in the machine vision (research) community, this verification is seldom done adequately.
暂无评论