In order to solve the problems of local maximum modulus extraction and threshold selection in the edge detection of finite resolution digital images, a new wavelet transform based adaptive dual threshold edge detec...In order to solve the problems of local maximum modulus extraction and threshold selection in the edge detection of finite resolution digital images, a new wavelet transform based adaptive dual threshold edge detection algorithm is proposed. The local maximum modulus is extracted by linear interpolation in wavelet domain. With the analysis on histogram, the image is filtered with an adaptive dual threshold method, which effectively detects the contours of small structures as well as the boundaries of large objects. A wavelet domain's propagation function is used to further select weak edges. Experimental results have shown the self adaptivity of the threshold to images having the same kind of histogram, and the efficiency even in noise tampered images.展开更多
A novel method of Doppler frequency extraction is proposed for Doppler radar scoring systems. The idea is that the time-frequency map can show how the Doppler frequency varies along the time-line, so the Doppler frequ...A novel method of Doppler frequency extraction is proposed for Doppler radar scoring systems. The idea is that the time-frequency map can show how the Doppler frequency varies along the time-line, so the Doppler frequency extraction becomes curve detection in the image-view. A set of morphological operations are used to implement curve detection. And a map fusion scheme is presented to eliminate the influence of strong direct current (DC) component of echo signal during curve detection. The radar real-life data are used to illustrate the performance of the new approach. Experimental results show that the proposed method can overcome the shortcomings of piecewise-processing-based FFT method and can improve the measuring precision of miss distance.展开更多
A novel image denoising method based on curvelet transform is proposed in order to improve the performance of Doppler frequency extraction in low signal-noise-ratio (SNR) environment. The echo can be represented as a ...A novel image denoising method based on curvelet transform is proposed in order to improve the performance of Doppler frequency extraction in low signal-noise-ratio (SNR) environment. The echo can be represented as a gray image with spectral intensity as its gray values by time-frequency transform. And the curvelet coefficients of the image are computed. Then an adaptive soft-threshold scheme based on dual-median operation is implemented in curvelet domain. After that, the image is reconstructed by inverse curvelet transform and the Doppler curve is extracted by a curve detection scheme. Experimental results show the proposed method can improve the detection of Doppler frequency in low SNR environment.展开更多
A resource allocation problem considering both efficiency and fairness in orthogonal frequency division multiple access (OFDMA) systems is studied. According to the optimality conditions, a downlink resource allocat...A resource allocation problem considering both efficiency and fairness in orthogonal frequency division multiple access (OFDMA) systems is studied. According to the optimality conditions, a downlink resource allocation algorithm consisting of subcarrier assignment and power alloca- tion is proposed. By adjusting the tradeoff coefficient, the proposed algorithm can achieve different levels of compromise between efficiency and fairness. The well-known classic resource allocation policies such as sum-rate maximization algorithm, proportional fairness algorithm and max-rain algorithm are all special cases of the proposed algorithm. Simulation results show that the compromise between efficiency and fairness can be continuously adjusted according to system requirements.展开更多
Image restoration is the problem of restoring a real degraded image.Previous studies mostly focused on single distortion.However,most of the real images experience multiple distortions,and single distortion image rest...Image restoration is the problem of restoring a real degraded image.Previous studies mostly focused on single distortion.However,most of the real images experience multiple distortions,and single distortion image restoration algorithms can not effectively improve the image quality.Moreover,few existing hybrid distortion image restoration algorithms can not deal with single distortion.Therefore,an end-to-end pipeline network based on stagewise training is proposed in this paper.Specifically,the network selects three typical image restoration tasks:denoising,inpainting,and super resolution.The whole training process is divided into single distortion training,hybrid distortion training of two types,and hybrid distortion training of three types.The design of loss function draws on the idea of deep supervision.Experimental results prove that the proposed method is not only superior to other methods in hybrid-distorted image restoration,but also suitable for single distortion image restoration.展开更多
文摘In order to solve the problems of local maximum modulus extraction and threshold selection in the edge detection of finite resolution digital images, a new wavelet transform based adaptive dual threshold edge detection algorithm is proposed. The local maximum modulus is extracted by linear interpolation in wavelet domain. With the analysis on histogram, the image is filtered with an adaptive dual threshold method, which effectively detects the contours of small structures as well as the boundaries of large objects. A wavelet domain's propagation function is used to further select weak edges. Experimental results have shown the self adaptivity of the threshold to images having the same kind of histogram, and the efficiency even in noise tampered images.
基金the Ministerial Level Advanced Research Foundation(020045089)
文摘A novel method of Doppler frequency extraction is proposed for Doppler radar scoring systems. The idea is that the time-frequency map can show how the Doppler frequency varies along the time-line, so the Doppler frequency extraction becomes curve detection in the image-view. A set of morphological operations are used to implement curve detection. And a map fusion scheme is presented to eliminate the influence of strong direct current (DC) component of echo signal during curve detection. The radar real-life data are used to illustrate the performance of the new approach. Experimental results show that the proposed method can overcome the shortcomings of piecewise-processing-based FFT method and can improve the measuring precision of miss distance.
文摘A novel image denoising method based on curvelet transform is proposed in order to improve the performance of Doppler frequency extraction in low signal-noise-ratio (SNR) environment. The echo can be represented as a gray image with spectral intensity as its gray values by time-frequency transform. And the curvelet coefficients of the image are computed. Then an adaptive soft-threshold scheme based on dual-median operation is implemented in curvelet domain. After that, the image is reconstructed by inverse curvelet transform and the Doppler curve is extracted by a curve detection scheme. Experimental results show the proposed method can improve the detection of Doppler frequency in low SNR environment.
文摘A resource allocation problem considering both efficiency and fairness in orthogonal frequency division multiple access (OFDMA) systems is studied. According to the optimality conditions, a downlink resource allocation algorithm consisting of subcarrier assignment and power alloca- tion is proposed. By adjusting the tradeoff coefficient, the proposed algorithm can achieve different levels of compromise between efficiency and fairness. The well-known classic resource allocation policies such as sum-rate maximization algorithm, proportional fairness algorithm and max-rain algorithm are all special cases of the proposed algorithm. Simulation results show that the compromise between efficiency and fairness can be continuously adjusted according to system requirements.
文摘Image restoration is the problem of restoring a real degraded image.Previous studies mostly focused on single distortion.However,most of the real images experience multiple distortions,and single distortion image restoration algorithms can not effectively improve the image quality.Moreover,few existing hybrid distortion image restoration algorithms can not deal with single distortion.Therefore,an end-to-end pipeline network based on stagewise training is proposed in this paper.Specifically,the network selects three typical image restoration tasks:denoising,inpainting,and super resolution.The whole training process is divided into single distortion training,hybrid distortion training of two types,and hybrid distortion training of three types.The design of loss function draws on the idea of deep supervision.Experimental results prove that the proposed method is not only superior to other methods in hybrid-distorted image restoration,but also suitable for single distortion image restoration.