Aiming at the problems of image super-resolution algorithm with many convolutional neural networks, such as large parameters, large computational complexity and blurred image texture, we propose a new algorithm model....Aiming at the problems of image super-resolution algorithm with many convolutional neural networks, such as large parameters, large computational complexity and blurred image texture, we propose a new algorithm model. The classical convolutional neural network is improved, the convolution kernel size is adjusted, and the parameters are reduced;the pooling layer is added to reduce the dimension. Reduced computational complexity, increased learning rate, and reduced training time. The iterative back-projection algorithm is combined with the convolutional neural network to create a new algorithm model. The experimental results show that compared with the traditional facial illusion method, the proposed method can obtain better performance.展开更多
Image super-resolution (SR) reconstruction is to reconstruct a high-resolution (HR) image from one or a series of low-resolution (LR) images in the same scene with a certain amount of prior knowledge. Learning based a...Image super-resolution (SR) reconstruction is to reconstruct a high-resolution (HR) image from one or a series of low-resolution (LR) images in the same scene with a certain amount of prior knowledge. Learning based algorithm is an effective one in image super-resolution reconstruction algorithm. The core idea of the algorithm is to use the training examples of image to increase the high frequency information of the test image to achieve the purpose of image super-resolution reconstruction. This paper presents a novel algorithm for image super resolution based on morphological component analysis (MCA) and dictionary learning. The MCA decomposition based SR algorithm utilizes MCA to decompose an image into the texture part and the structure part and only takes the texture part to train the dictionary. The reconstruction of the texture part is based on sparse representation, while that of the structure part is based on more faster method, the bicubic interpolation. The proposed method improves the robustness of the image, while for different characteristics of textures and structure parts, using a different reconstruction algorithm, better preserves image details, improve the quality of the reconstructed image.展开更多
Facing the body's EEG(electroencephalograph, 0.5–100 Hz, 5–100 μV) and ECG's(electrocardiogram,〈 100 Hz, 0.01–5 mV) micro signal detection requirement, this paper develops a pervasive application micro sign...Facing the body's EEG(electroencephalograph, 0.5–100 Hz, 5–100 μV) and ECG's(electrocardiogram,〈 100 Hz, 0.01–5 mV) micro signal detection requirement, this paper develops a pervasive application micro signal detection ASIC chip with the chopping modulation/demodulation method. The chopper-stabilization circuit with the RRL(ripple reduction loop) circuit is to suppress the ripple voltage, which locates at the single-stage amplifier's outputting terminal. The single-stage chopping core's noise has been suppressed too, and it is beneficial for suppressing noises of post-circuit. The chopping core circuit uses the PFB(positive feedback loop) to increase the inputting resistance, and the NFB(negative feedback loop) to stabilize the 40 dB intermediate frequency gain. The cascaded switch-capacitor sample/hold circuit has been used for deleting spike noises caused by non-ideal MOS switches, and the VGA/BPF(voltage gain amplifier/band pass filter) circuit is used to tune the chopper system's gain/bandwidth digitally. Assisted with the designed novel dry-electrode, the real test result of the chopping amplifying circuit gives some critical parameters: 8.1 μW/channel, 0.8 μVrms(@band-widthD100 Hz), 4216–11220 times digitally tuning gain range, etc. The data capture system uses the NI CO's data capturing DAQmx interface,and the captured micro EEG/ECG's waves are real-time displayed with the PC-Labview. The proposed chopper system is a unified EEG/ECG signal's detection instrument and has a critical real application value.展开更多
文摘Aiming at the problems of image super-resolution algorithm with many convolutional neural networks, such as large parameters, large computational complexity and blurred image texture, we propose a new algorithm model. The classical convolutional neural network is improved, the convolution kernel size is adjusted, and the parameters are reduced;the pooling layer is added to reduce the dimension. Reduced computational complexity, increased learning rate, and reduced training time. The iterative back-projection algorithm is combined with the convolutional neural network to create a new algorithm model. The experimental results show that compared with the traditional facial illusion method, the proposed method can obtain better performance.
文摘Image super-resolution (SR) reconstruction is to reconstruct a high-resolution (HR) image from one or a series of low-resolution (LR) images in the same scene with a certain amount of prior knowledge. Learning based algorithm is an effective one in image super-resolution reconstruction algorithm. The core idea of the algorithm is to use the training examples of image to increase the high frequency information of the test image to achieve the purpose of image super-resolution reconstruction. This paper presents a novel algorithm for image super resolution based on morphological component analysis (MCA) and dictionary learning. The MCA decomposition based SR algorithm utilizes MCA to decompose an image into the texture part and the structure part and only takes the texture part to train the dictionary. The reconstruction of the texture part is based on sparse representation, while that of the structure part is based on more faster method, the bicubic interpolation. The proposed method improves the robustness of the image, while for different characteristics of textures and structure parts, using a different reconstruction algorithm, better preserves image details, improve the quality of the reconstructed image.
基金Project supported by the National Natural Science Foundation of China(Nos.61527815,31500800,61501426,61471342)the National Key Basic Research Plan(No.2014CB744600)+1 种基金the Beijing Science and Technology Plan(No.Z141100000214002)the Chinese Academy of Sciences’Key Project(No.KJZD-EW-L11-2)
文摘Facing the body's EEG(electroencephalograph, 0.5–100 Hz, 5–100 μV) and ECG's(electrocardiogram,〈 100 Hz, 0.01–5 mV) micro signal detection requirement, this paper develops a pervasive application micro signal detection ASIC chip with the chopping modulation/demodulation method. The chopper-stabilization circuit with the RRL(ripple reduction loop) circuit is to suppress the ripple voltage, which locates at the single-stage amplifier's outputting terminal. The single-stage chopping core's noise has been suppressed too, and it is beneficial for suppressing noises of post-circuit. The chopping core circuit uses the PFB(positive feedback loop) to increase the inputting resistance, and the NFB(negative feedback loop) to stabilize the 40 dB intermediate frequency gain. The cascaded switch-capacitor sample/hold circuit has been used for deleting spike noises caused by non-ideal MOS switches, and the VGA/BPF(voltage gain amplifier/band pass filter) circuit is used to tune the chopper system's gain/bandwidth digitally. Assisted with the designed novel dry-electrode, the real test result of the chopping amplifying circuit gives some critical parameters: 8.1 μW/channel, 0.8 μVrms(@band-widthD100 Hz), 4216–11220 times digitally tuning gain range, etc. The data capture system uses the NI CO's data capturing DAQmx interface,and the captured micro EEG/ECG's waves are real-time displayed with the PC-Labview. The proposed chopper system is a unified EEG/ECG signal's detection instrument and has a critical real application value.