The wavelet maxima representation is a new reorganization of an image information in order to solve the image processing problem from the properties of the image edges. The multiscale method of wavelet for edge detect...
详细信息
The wavelet maxima representation is a new reorganization of an image information in order to solve the image processing problem from the properties of the image edges. The multiscale method of wavelet for edge detection and reconstruction is presented. The experiment shows that image approximation may be reconstructed from the wavelet modulus maxima without the visual degradation.
The recognition method of the harbor target is presented. There are two aspects in the recognition. One is the recognition of the harbor contour, the other is the recognition of the texture of the harbor area on one s...
详细信息
The recognition method of the harbor target is presented. There are two aspects in the recognition. One is the recognition of the harbor contour, the other is the recognition of the texture of the harbor area on one side of the contour. The chain code of the contour represents the orientation and the curvature of the contour, and the control points are detected by the extrema of the wavelet transform after adaptive smoothing. The primitive of the harbor contour is some structure of successive control points. The primitive is recognized by three features related to these points, which are scale-and rotation-invariant.
It shows that prior information on the location where quantized value of large coefficients is comparative to that of coefficients is available in coding when natural images are decomposed on wavelet bases. The high e...
详细信息
It shows that prior information on the location where quantized value of large coefficients is comparative to that of coefficients is available in coding when natural images are decomposed on wavelet bases. The high efficiency of embedded coding is implemented. Wavelet embedded coding and implementation of zero-tree coding are also considered. The significant map of embedded wavelet coding is stored in zero-tree. A quad-tree is used to denote and significant map in wavelet zero-tree coding. Experiments and analysis show that images with high quality are obtained in the case of only 0.25bit/pixel. The computed and estimated peak to peak SNR cures are dovetailed excellently in wavelet embedded coding.
Based on the improvement of the processing of IC fabrication, an advanced JTAG controller is presented. The controller can address 16 MB memory frame sequentially and provides a better workbench for test soft programm...
详细信息
Based on the improvement of the processing of IC fabrication, an advanced JTAG controller is presented. The controller can address 16 MB memory frame sequentially and provides a better workbench for test soft programmer. A new implementation of the logic and circuit is carried out. The simulation proves that the chip can provide the test signals meeting IEEE P1149.1 protocol.
The method of wavelet thresholding based on image denoising is presented. The image contaminated by a white noise is decomposed in a separable two-dimensional wavelet basis. A wavelet thresholding algorithm which is w...
详细信息
The method of wavelet thresholding based on image denoising is presented. The image contaminated by a white noise is decomposed in a separable two-dimensional wavelet basis. A wavelet thresholding algorithm which is well adapted to suppressing noise from non-textured image is introduced to produce a better image quality with a higher SNR.
A method by using top-hat operator for infrared small target detection is proposed. The sequence images are filtered by morphologic operator;the filtered images: are segmented;segmented images are continuously filtere...
详细信息
A method by using top-hat operator for infrared small target detection is proposed. The sequence images are filtered by morphologic operator;the filtered images: are segmented;segmented images are continuously filtered by depending on the feature of target's movement. The method is verified by the experiments.
According to the characteristics of column-linear-scan image, a new image processing method called column processing technology is presented. For conventional image processing approached, an image will not be processe...
详细信息
According to the characteristics of column-linear-scan image, a new image processing method called column processing technology is presented. For conventional image processing approached, an image will not be processed until its all data are got ready, while our method will handle each column of an image when it comes. The main advantages of the proposed method are that it has ability to decrease time requirement and improve the real-time performance of image systems. The experimental result of a target detection algorithm based on mathematical morphology on a DSP system proves the effectiveness of our method.
This paper presents an approach to realize finite precision chaotic systems by constructing a variable parameter compound chaotic system. By setting parameter list and switch condition, a discrete chaotic system can b...
详细信息
This paper presents an approach to realize finite precision chaotic systems by constructing a variable parameter compound chaotic system. By setting parameter list and switch condition, a discrete chaotic system can be transformed into the variable parameter chaotic system. Based on this, the variable parameter compound chaotic system can be constructed by coupling two or more than two variable parameter chaotic systems. How to set up a variable parameter chaotic system with a logistic map is demonstrated. The way to establish a variable parameter compound chaotic system is illustrated by compounding a parabola map and a tent map. It is shown that a great number of chaotic sequences with wonderful correlation properties can be generated conveniently by this method. Compared with the original chaotic systems, the sequences generated by the compound system derived from them have more smooth power spectra and larger lyapunov index.
Through the analysis of the features of bridge object in long-range IR image, a new recognition algorithm based on knowledge framework is put forward. Due to some characters of long-range infrared bridge images, the g...
详细信息
Through the analysis of the features of bridge object in long-range IR image, a new recognition algorithm based on knowledge framework is put forward. Due to some characters of long-range infrared bridge images, the global method processing the whole image equally is ineffective. Based on the features of the bridge object, our proposed algorithm first detects the candidate bridge regions and then focuses on them. According to the degree to which they match our prebuilding framework, different credits are given, so the false targets are excluded and eventually the real target is found. The experiment results demonstrate the good performance of our method.
A method for fractal coding for infrared image based on adaptive matching is presented. By using the correlation between image blocks, this method limits the matching between the range blocks and domain blocks to be o...
详细信息
A method for fractal coding for infrared image based on adaptive matching is presented. By using the correlation between image blocks, this method limits the matching between the range blocks and domain blocks to be on the dynamic windows centered at the mean of the standard deviation of the range blocks. The length of the windows and their width are adaptively adjusted by the planar histograms of the range blocks. By comparison with the conventional methods the experiments show that the approach can not only reduce the computation cost greatly but also increase the compression ratio efficiently.
暂无评论