31results about How to "Achieving Super-Resolution Reconstruction" patented technology

The invention discloses an image super-resolution reconstruction method based on a multi-core gaussian process regression and mainly solves the problems that the current super-resolution reconstruction method generates edge sawtooth effect and the reconstruction texture is not rich. The image super-resolution reconstruction method based on the multi-core gaussian process regression comprises the following steps: (1), obtaining a low-resolution luminance image and an interpolation image and blocking the low-resolution luminance image and the interpolation image; (2), extracting central pixels and eight neighborhoods of low-resolution luminance image blocks to train an upper sampling model of the gaussian process regression; (3), forecasting pixel values of initial high-resolution luminance image blocks by using the upper sampling model; (4), combining all the initial high-resolution luminance image blocks to obtain an initial high-resolution luminance image; (5), obtaining an analog low-resolution image and blocking the analog low-resolution image; (6), extracting central pixels of the analog low-resolution image blocks to train a deblurring model of the gaussian process regression; (7), forecasting pixel values of the high-resolution luminance image blocks by using the deblurring model; and (8), combining all the high-resolution luminance image blocks to obtain a high-resolution luminance image. The image super-resolution reconstruction method based on the multi-core gaussian process regression is applicable to video monitoring and imaging of high-definition televisions.

The invention provides a high-efficiency super-resolution imaging device and method with regional management, and aims at solving the problem that an existing super-resolution imaging device and method is long in imaging time and low in resolution. The imaging device comprises an imaging lens group, a high-resolution detector, an image storage module, an image pre-processing module, an image super-resolution reconstruction module and an image output display module connected successively, wherein the image super-resolution reconstruction module is composed of an image regional management sub-module and a dictionary training and regional reconstruction sub-module. The imaging method comprises the following steps of obtaining optical signals of a practical scene; obtaining low-resolution images; storing the low-resolution images; preprocessing the images; carrying out regional image management; training dictionaries and reconstructing super-resolution areas, and splicing images of the reconstructed super-resolution sub-regions. The imaging device and method can be used to improve the imaging resolution effectively and shorten the imaging time, and applied to the fields including video monitoring, satellite remote-sensing imaging and medical imaging.

The present invention relates to a people and certificate verification method and device. The method comprises a step of obtaining a certificate face image in a certificate picture and collecting a natural light face image of a user, a step of inputting the certificate face image into a pre-trained generative adversarial network, and obtaining a reconstruction face image corresponding to the certificate face image according to the output of the generative adversarial network, wherein the generative adversarial network is used for adding preset natural light attribute information to the inputted certificate face image, the resolution of the outputted reconstruction face image is higher than that of the certificate face image, and a step of comparing the reconstruction face image and the natural light face image and carrying out people and certificate verification according to a comparison result. According to the method and the device, the accuracy of people and certificate verification can be effectively improved.

The invention relates to a single-image super-resolution method based on a simplified ESRGAN, and the method comprises the following steps: S1, obtaining a to-be-processed low-resolution image, and carrying out the preprocessing of the to-be-processed low-resolution image; s2, according to the preprocessed image, generating a super-resolution image through a generator module in the improved single-image super-resolution generative adversarial network, if the model is in a training stage, carrying out the step S3, and otherwise, carrying out the step S4; s3, constructing a discriminator, usingthe discriminator to judge whether the super-resolution image is a real high-resolution image or not, performing back propagation according to a result obtained by the discriminator, optimizing the generator, and performing the step S2 again; and S4, carrying out edge restoration processing on the obtained super-resolution image to obtain a final super-resolution image. According to the method, the problem of edge restoration after image amplification is solved, the edge sawtooth effect and the blocking effect are removed, the image is smoother, and therefore single-image super-resolution reconstruction is well achieved.

The invention discloses a wavelet preprocessing and sparse representation-based satellite remote sensing image super-resolution reconstruction method and belongs to the technical field of satellite remote sensing image processing. The wavelet preprocessing and sparse representation-based satellite remote sensing image super-resolution reconstruction method is applicable to high resolution remote sensing images and low resolution remote sensing images with different time resolutions in the same known observation area, super-resolution reconstruction is performed on low resolution remote sensing images at other observation times and the spatial resolution of the low resolution remote sensing images is improved. The wavelet preprocessing and sparse representation-based satellite remote sensing image super-resolution reconstruction method specifically comprises steps of dictionary training and low resolution remote sensing image reconstruction. According to the wavelet preprocessing and sparse representation-based satellite remote sensing image super-resolution reconstruction method, the phonological change of the remote sensing image is taken into consideration, wavelet domain dictionaries comprising different character information are constructed, super-resolution reconstruction of the low resolution remote sensing images is effectively achieved based on training of the three pairs of wavelet section dictionaries and in combination with sparse representation, image detail features are well obtained, the reconstruction quality of the low resolution remote sensing images is effectively improved, and a basis is provided for later applications of the low resolution remote sensing images.

The invention discloses a stereo image super-resolution reconstruction method based on deep interactive learning, and the method comprises the steps: dividing an input left view and an input right view into a left branch and a right branch, and extracting corresponding spatial feature expressions through spatial features; extracting complementary information in another viewpoint through an interaction part and adopting the complementary information for enhancing spatial feature expression of the left view and the right view; adopting a mean square error loss function, a gradient loss functionand a parallax loss function for jointly constructing a multi-loss function mechanism, and adopting the multi-loss function mechanism for improving stereo image super-resolution reconstruction quality; and training a stereo image super-resolution reconstruction network based on deep interactive learning. According to the method, the spatial correlation and the inter-viewpoint correlation of the left view and the right view are obtained by mining complementary information in the stereo image by utilizing the feature expression capability of deep learning.

The invention discloses a single image super-resolution reconstruction method based on a lightweight neural network and a Transform. Firstly, a low-frequency feature extraction module is used for extracting a low-frequency feature map of a low-resolution image, secondly, the feature map is input into a frame formed by combining a depth separable convolution module and a Transform in a blocking mode, high-frequency texture detail features of the image are extracted, and finally, the low-frequency feature map and the high-frequency feature map are subjected to jump connection to output a high-resolution image. According to the method, super-resolution reconstruction of the low-resolution image can be realized, and the high-resolution image with rich details, clear texture and high spatial resolution is obtained.

The invention provides a portrait super-resolution reconstruction method and device, a model training method and device, electronic equipment and a readable storage medium, and the method comprises the steps: carrying out the key point detection of a to-be-processed image through a pre-constructed reconstruction model, and obtaining face key points; and performing super-resolution reconstruction processing according to the face key points and image features obtained based on the to-be-processed image to obtain image high-frequency information, and performing restoration processing on the to-be-processed image by using the image high-frequency information to obtain a super-resolution image corresponding to the to-be-processed image. According to the method and the device, the face key point detection and the face recovery are combined, the super-resolution reconstruction of the image is realized, the cognition degree of the obtained super-resolution image is improved, and the requirement of a user in practical application is met.

The invention discloses an unsupervised remote sensing image super-resolution reconstruction method based on image recursion. Obtaining an original low-resolution image; extracting a degradation kernel from the original low-resolution image; performing down-sampling on the original low-resolution image by using a degradation kernel to obtain a low-resolution image after down-sampling; training the super-resolution reconstruction network model based on the original low-resolution image and the down-sampled low-resolution image; and adopting the super-resolution reconstruction network model to test the to-be-tested low-resolution image to obtain a corresponding super-resolution reconstruction result. Through the method, a super-resolution reconstruction method can be applied to the remote sensing image, so that the accuracy of obtaining pairing data from the remote sensing image is improved, and a network realizes a good reconstruction effect on downsampling used for reconstruction synthesis.

The invention discloses a three-dimensional image super-resolution reconstruction method based on cyclic interaction, which comprises the following steps: recombining multilayer features of left and right viewpoints into left and right sequences through queue recombination conversion, and recombining arrangement following the sequence of the features from a shallow layer to a deep layer; building a circular interaction module, enhancing multi-layer features of left and right viewpoints interactively through a circular structure, wherein the circular interaction module is composed of a circular interaction unit, and the circular interaction unit is composed of two interaction units and a jump connection; through a multi-propagation strategy, circularly interactively inputting multilayer features of left and right viewpoints in a sequence, learning dependency between viewpoints to enhance the features, and further obtaining final circular interaction enhanced features; enhancing features based on circulation interaction, using sub-pixel convolution to improve feature resolution, and using n * n convolution to reconstruct the features into high-resolution left and right views; and building a multi-loss function mechanism by using a correlation loss function, a difference loss function and an L1 loss function, so that the super-resolution reconstruction quality of the three-dimensional image is improved.