In this study,an underwater image enhancement method based on multi-scale adversarial network was proposed to solve the problem of detail blur and color distortion in underwater images.Firstly,the local features of ea...In this study,an underwater image enhancement method based on multi-scale adversarial network was proposed to solve the problem of detail blur and color distortion in underwater images.Firstly,the local features of each layer were enhanced into the global features by the proposed residual dense block,which ensured that the generated images retain more details.Secondly,a multi-scale structure was adopted to extract multi-scale semantic features of the original images.Finally,the features obtained from the dual channels were fused by an adaptive fusion module to further optimize the features.The discriminant network adopted the structure of the Markov discriminator.In addition,by constructing mean square error,structural similarity,and perceived color loss function,the generated image is consistent with the reference image in structure,color,and content.The experimental results showed that the enhanced underwater image deblurring effect of the proposed algorithm was good and the problem of underwater image color bias was effectively improved.In both subjective and objective evaluation indexes,the experimental results of the proposed algorithm are better than those of the comparison algorithm.展开更多
In order to extract usable harmonics from real 2^(n) sequence pseudo-random data,a technical method is proposed.An equation for predicting the average amplitude of the main frequencies is proposed to guide the choice ...In order to extract usable harmonics from real 2^(n) sequence pseudo-random data,a technical method is proposed.An equation for predicting the average amplitude of the main frequencies is proposed to guide the choice of signal type for different exploration tasks.By the threshold of the amplitude of the transmitted signal,a set of candidate frequencies are first selected.Then,by operating a spectrum envelope method at these candidate frequencies on received data,effective components in data are extracted.A frequency density calculation method is proposed based on a logical number summation method,to reasonably characterize the frequency density in different frequency bands.By applying this method to real data in Sichuan,China,with signal Type 13,75 effective components are extracted,including both main frequencies and harmonics.The result suggests that the number of effective frequencies in the 2^(n) sequence pseudo-random signal can be increased by extracting usable harmonics,without any additional fieldwork.展开更多
Different cell populations from bone marrow are used in various clinical trials for cardiac diseases during last decade. Four clinical studies are on going in our institution and enrol patients with cardiac diseases, ...Different cell populations from bone marrow are used in various clinical trials for cardiac diseases during last decade. Four clinical studies are on going in our institution and enrol patients with cardiac diseases, coronary disease and type 2 diabetes, patients with osteoarthritis. Density gradient is used to separate bone marrow mononuclear cells. Cell processing looses are significant. To find out critical control points we analysed processing process and differences in cell yields between operators performing cell extraction. Bone marrow mononuclear cells were isolated using Ficoll density gradient centrifugation. Cells were counted using flow cytometry for mononuclear cell total counts, CD34＋ population count and viability analysis. The patients who underwent bone marrow aspiration followed by cell isolation received cell suspension for transplantation. Two cells processing for separate patients were performed at once. Same standard operation procedures were applied. Processing looses between operators performing cell extraction were analysed. Bone marrow samples from eight patients were processed. Mononuclear cells were extracted. Operator performances were compared. Operator A average bone marrow mononuclear cell yield in starting material was 9,97 ± 9,98 %, CD34＋ population yield 75,46±79,67%. Operator B average bone marrow mononuclear cell yield in starting material was 24,68 ± 14,8 %, CD34＋ population yield 70,42 ±44.84%. Operator A average cell viability in starting material was 45,24 ± 9,55%, after cell processing 42,96 ± 23,66 %. Operator B average cell viability in starting material was 49,85 ± 5,48%, after cell processing 69,52 ± 6,65 %.展开更多
文摘In this study,an underwater image enhancement method based on multi-scale adversarial network was proposed to solve the problem of detail blur and color distortion in underwater images.Firstly,the local features of each layer were enhanced into the global features by the proposed residual dense block,which ensured that the generated images retain more details.Secondly,a multi-scale structure was adopted to extract multi-scale semantic features of the original images.Finally,the features obtained from the dual channels were fused by an adaptive fusion module to further optimize the features.The discriminant network adopted the structure of the Markov discriminator.In addition,by constructing mean square error,structural similarity,and perceived color loss function,the generated image is consistent with the reference image in structure,color,and content.The experimental results showed that the enhanced underwater image deblurring effect of the proposed algorithm was good and the problem of underwater image color bias was effectively improved.In both subjective and objective evaluation indexes,the experimental results of the proposed algorithm are better than those of the comparison algorithm.
基金financially supported by the National Key Research and Development Program of China(No.2019YFC0604902)the National Natural Science Foundation of China(No.42004056)the Natural Science Foundation of Shandong Province,China(No.ZR201911010111).
文摘In order to extract usable harmonics from real 2^(n) sequence pseudo-random data,a technical method is proposed.An equation for predicting the average amplitude of the main frequencies is proposed to guide the choice of signal type for different exploration tasks.By the threshold of the amplitude of the transmitted signal,a set of candidate frequencies are first selected.Then,by operating a spectrum envelope method at these candidate frequencies on received data,effective components in data are extracted.A frequency density calculation method is proposed based on a logical number summation method,to reasonably characterize the frequency density in different frequency bands.By applying this method to real data in Sichuan,China,with signal Type 13,75 effective components are extracted,including both main frequencies and harmonics.The result suggests that the number of effective frequencies in the 2^(n) sequence pseudo-random signal can be increased by extracting usable harmonics,without any additional fieldwork.
文摘Different cell populations from bone marrow are used in various clinical trials for cardiac diseases during last decade. Four clinical studies are on going in our institution and enrol patients with cardiac diseases, coronary disease and type 2 diabetes, patients with osteoarthritis. Density gradient is used to separate bone marrow mononuclear cells. Cell processing looses are significant. To find out critical control points we analysed processing process and differences in cell yields between operators performing cell extraction. Bone marrow mononuclear cells were isolated using Ficoll density gradient centrifugation. Cells were counted using flow cytometry for mononuclear cell total counts, CD34＋ population count and viability analysis. The patients who underwent bone marrow aspiration followed by cell isolation received cell suspension for transplantation. Two cells processing for separate patients were performed at once. Same standard operation procedures were applied. Processing looses between operators performing cell extraction were analysed. Bone marrow samples from eight patients were processed. Mononuclear cells were extracted. Operator performances were compared. Operator A average bone marrow mononuclear cell yield in starting material was 9,97 ± 9,98 %, CD34＋ population yield 75,46±79,67%. Operator B average bone marrow mononuclear cell yield in starting material was 24,68 ± 14,8 %, CD34＋ population yield 70,42 ±44.84%. Operator A average cell viability in starting material was 45,24 ± 9,55%, after cell processing 42,96 ± 23,66 %. Operator B average cell viability in starting material was 49,85 ± 5,48%, after cell processing 69,52 ± 6,65 %.
