A function that can distinguish various edge patterns of the input vectors in the frequency domain is proposed, and a classifier is designed using a cluster-seeking algorithm to ensure that the weighted mean-square er...
详细信息
A function that can distinguish various edge patterns of the input vectors in the frequency domain is proposed, and a classifier is designed using a cluster-seeking algorithm to ensure that the weighted mean-square error is minimum for each equal size class. For a given image training sequence, it is shown that this approach can classify the input vectors very effectively and makes the classifier design very simple. Good visual quality was obtained with 4 multiplied by 4 blocks at a rate in the range of 0. 56-0. 625 b/pixel, while encoding complexity was far less than that of other methods.
The authors have developed digital captioning devices using available word processors. These devices are called video information processors (VIP). The first, the VIP-2000, was developed in 1983. However, the VIP-2000...
详细信息
ISBN:
(纸本)4274031888
The authors have developed digital captioning devices using available word processors. These devices are called video information processors (VIP). The first, the VIP-2000, was developed in 1983. However, the VIP-2000 was in the pattern format and sometimes too slow to match captions with visual speech. Therefore, the authors recently developed the VIP-4000 in the code format. This device is fast enough to match captions with visual speech. The VIP-4000 also has other available functions. For example, both the horizontal and vertical edges can be added, and one can color character by character. Compared with the usual method, the authors' system succeed in reducing greatly the time required to caption video tapes and remain almost faithful to speech.
A new segmentation method based on the properties of the human visual system is described in this paper. The segmentation method is part of a new second generation image coder. In addition, one of the properties is th...
详细信息
A new segmentation method based on the properties of the human visual system is described in this paper. The segmentation method is part of a new second generation image coder. In addition, one of the properties is that it is not necessary to transmit (or store) the visual residual for use in reconstructing the received signal. It is assumed that the characteristics of this visual residual are known at the receiver and can be used in the reconstruction process.
Recently, ways to obtain a new generation of image-coding techniques have been proposed. The incorporation of the human visual system (HVS) models and tools of the image analysis, such as segmentation, are two definin...
详细信息
Recently, ways to obtain a new generation of image-coding techniques have been proposed. The incorporation of the human visual system (HVS) models and tools of the image analysis, such as segmentation, are two defining features of these techniques. In this paper, a new image coding technique is presented which utilizes a human visual system model and a image analysis criteria together with the technique of projection onto convex sets (POCS). The human visual system model is used to determine the most probable perceptual information of an image and the removal of irrelevant information. The projection onto convex sets operation is used as a restoration technique aimed at correcting the degradations caused by the conventional coders in the transmitter part of the channel. And finally, the image analysis based technique is used as an interpolator in order to predict the unknown regions of the received image. This new technique can be used in single frame image coding applications to obtain acceptable images at extremely high compression rates.
image understanding requires the use of specifications of object structure and appearance in terms of extractable image properties. A description is given of experimentation with a schemata-based approach to visual in...
详细信息
ISBN:
(纸本)0818606916
image understanding requires the use of specifications of object structure and appearance in terms of extractable image properties. A description is given of experimentation with a schemata-based approach to visual interpretation which requires the construction of a recognition network that is subsequently coded in the programming language OCCAM to obtain a full expression of the parallelism inherent in the interpretation process.
A method which combines vector quantization (VQ) and orthogonal transform coding schemes is presented. The use of these two techniques in image coding allows efficient bit-rate reduction (about 8-10) with an original-...
详细信息
A method which combines vector quantization (VQ) and orthogonal transform coding schemes is presented. The use of these two techniques in image coding allows efficient bit-rate reduction (about 8-10) with an original-like reconstruction quality and easy introduction of visual properties in both classification procedure and codebook design. Recent work, including a classification algorithm and transformed coefficients coding, is presented. image sequence coding using VQ is then discussed, and some strategies for codebook refreshment are proposed. 7 refs.
An approach to image coding is presented that is based on the theory of fractal geometry. The idea is to traverse the entire data string utilizing a fixed length 'yardstick'. The coding is achieved by transmit...
详细信息
An approach to image coding is presented that is based on the theory of fractal geometry. The idea is to traverse the entire data string utilizing a fixed length 'yardstick'. The coding is achieved by transmitting only the sign bit (to distinguish between the ascent and the descent) and the horizontal distance covered by the 'yardstick'. All data values are estimated at the receivers site, on the basis of this information. The approach can be utilized for classification of the image complexity and can be used directly for the image compression. The tradeoff between the resolution and quantization errors is easily controlled and a good utilization of human visual characteristics is achieved. Even at the present preliminary stage, excellent quality images have been obtained with as few as 0. 5 b/pel.
作者：
Martinex, K.Pearson, D.E.Univ of Essex
Dep of Electrical Engineering Science Colchester Engl Univ of Essex Dep of Electrical Engineering Science Colchester Engl
Consideration is given to the trade-off between algorithmic complexity and processing speed in primitive extraction for low data-rate visual communication. Initial experiments with serially-implemented algorithms on a...
详细信息
ISBN:
(纸本)0903748592
Consideration is given to the trade-off between algorithmic complexity and processing speed in primitive extraction for low data-rate visual communication. Initial experiments with serially-implemented algorithms on an MC68000 processor indicated the need for a high degree of parallelism in order to meet the requirements of real-time, moving-picture data rates. However, insight gained by trials on CLIP4, which typically showed speed gains of 40%, points to window-based architectures implemented using look-up tables.
The bandwidth compression of images through the use of a transform source encoding algorithm is described in general terms. The key element of the algorithm, the quantization procedure, is then described in detail. Th...
详细信息
The bandwidth compression of images through the use of a transform source encoding algorithm is described in general terms. The key element of the algorithm, the quantization procedure, is then described in detail. The development of optimal quantizers with an entropy constraint and with a distortion criterion (either mean square error or visual) is then described. Finally, the use of a different linear quantizer for each DCT frequency coefficient is presented in detail.
The following topics are dealt with: expert systems;computer vision;control;biocybernetics;communications, speech, and signal processing;fuzzy information;visual system;pattern recognition;identification and estimatio...
详细信息
The following topics are dealt with: expert systems;computer vision;control;biocybernetics;communications, speech, and signal processing;fuzzy information;visual system;pattern recognition;identification and estimation;systems engineering;decision-making;large-scale systems;adaptive and learning systems;robotics;simulation and modeling;management;imageprocessing;man-machine systems;computer systems;database systems;risk and fault analysis;public systems and policy analysis;remote sensing;energy systems;socioeconomic systems and technologies for developing countries;international stability;and optimization and systems method. 360 papers were presented, of which 276 are published in full in the present proceedings, and 13 as abstracts only. *
暂无评论