36 results about "Hyperspectral image fusion" patented technology

The invention relates to a wavelet transformation and multi-channel PCNN-based hyperspectral image fusion method, which comprises the following steps: firstly, performing preprocessing of registering and grey level adjustment on hyperspectral images of N wave bands to be fused, and performing the wavelet transformation to obtain low-frequency sub-band images and high-frequency sub-band images; secondly, performing primary nonlinear fusion processing on the low-frequency sub-band images and the high-frequency sub-band images respectively by using a multi-channel PCNN model, obtaining corresponding ignition frequency map, performing linear mapping of corresponding coefficient range on the ignition frequency map for the low-frequency sub-band images, and taking a mapping result as a fusion result; thirdly, performing the region segmentation on the high-frequency sub-band images in each direction by using the ignition frequency map, and performing the fusion processing on different regions by using different fusion rules; and finally, processing wavelet reconstruction and obtaining a final result image. The method can realize the hyperspectral image fusion of a plurality of hyperspectral wave bands and can achieve a better fusion effect.

The invention provides a method for reconstructing a wide hyperspectral image based on fusion of mulspectral/hyperspectral images. The method for reconstructing the wide hyperspectral image based on fusion of the multispectral/hyperspectral images comprises the following steps that surface feature end members are synchronously extracted in an overlapping area of the multispectral/hyperspectral images; a fusion model among the multispectral/hyperspectral images is established according to the end members, a transformational relation is established, model parameters are resolved and calculated, and a model parameter base is established; selection of the model parameters is conducted through spectrum matching, and spectrum reconstruction is conducted on multispectral images pixels by pixels so that hyperspectral information can be obtained. According to the technical scheme, by means of data fusion, successive wide hyperspectral images which have high spectral resolutions can be obtained through reconstruction of other multispectral remote sensing data, the spectral resolutions of the hyperspectral images are identical to data of original hyperspectral images, the spatial resolution and the width are identical to the original multispectral data, the hyperspectral resolutions of the original hyperspectral images is kept, and the spatial resolution and the width of each of the hyperspectral images can be improved.

The invention belongs to the field of hyperspectral image fusion processing and specially relates to a hyperspectral image fusion method for hyperspectral image fusion and spatial resolution enhancement based on end member extraction and spectrum unmixing. The method comprises the steps of: using an N-FINDR algorithm to carry out end member extraction; using a spectrum unmixing technology to obtain an abundance value of each end member in each pixel; using an abundance matrix A as prior knowledge, carrying out classified marking on pixels of a plurality of spectral images by means of a fuzzy C mean value clustering algorithm, and then carrying out fused image reconstruction according to marking results and end member spectrums; obtaining classifying results, assigning end member spectrums to each pixel of a hyperspectral image according to marked categories, and obtaining a reconstructed fused hyperspectral image. According to the invention, the end member extraction technology is used for extracting and reserving end member spectrum information, no coefficient conversion steps are introduced in the whole fusion process, so that spectrum distortion is avoided; in addition, compared with an existing hyperspectral image fusion method, the hyperspectral image fusion method provided by the invention is better in spectrum fidelity.

The invention discloses a coupling tensor decomposition-based multispectral image and hyperspectral image fusion method, and mainly aims at solving the problems of space information fuzziness and spectrum distortion in the prior art. The method comprises the following steps of: (1) respectively inputting a low-spatial resolution hyperspectral image, a high-spatial resolution multispectral image, aspectrum degradation matrix and a regular parameter; (2) carrying out spatial up-sampling on the low-spatial resolution hyperspectral image; (3) carrying out Tucker decomposition on the up-sampled low-spatial resolution hyperspectral image by utilizing a high-order singular value decomposition algorithm; (4) carrying out Tucker decomposition on the high-spatial resolution multispectral image; and(5) multiplying a factor matrix obtained in the step (3) by a core tensor obtained in the step (4) along each model so as to obtain a high-spatial resolution hyperspectral image. The method is capable of obtaining correct space and spectrum information and can be applied to the remote sensing fields such as target detection and recognition and terrain classification.

The invention relates to a hyperspectral image fusion method based on spectrum restrain, which is a hyperspectral image and high spatial resolution fusion method based on a nonnegative matrix factorization method. The method comprises three steps, including step 1, a computer reads hyperspectral images under a matrix laboratory (MATLAB) 7.1 environment; step 2, a fusion model based on the nonnegative matrix factorization method is built; and step 3, extreme value functions of the fusion model are solved by using a gradient descent method, and fusion results are obtained. The hyperspectral image fusion method overcomes the shortcoming that the method in the prior art cannot be suitable to directly transplanted to hyperspectral image fusion, and is poor in spectrum retentivity, obtains good fusion results, and has wide application prospect in the technical field of the hyperspectral image fusion.

The present invention discloses a hyperspectral and multispectral image rapid fusion method. The method comprises the implementation steps of: (1) establishing a training set and a test set; (2) inputting a spectrum image; (3) establishing a probability generation model; (4) training a probability generation model; and (5) obtaining a fusion spectrum image. The hyperspectral and multispectral image rapid fusion method overcomes the problems that a representational capacity is weak, the model is not stable and the operation time is long in the prior art, generated hyperspectral image with highresolution comprises more information, a space transfer matrix without priori conditions is employed to allow the model to be more stable and allow the fusion speed to be faster, and therefore the hyperspectral and multispectral image rapid fusion method is a high-speed and efficient hyperspectral image fusion processing method.

The invention provides a high spectral image fusion method based on a variational method, which is a high spectral image and high-resolution visible light image fusion method based on a variational method. The fusion method comprises the following three steps: (1) reading a high spectral image data by a computer in an MATLAB7.1 environment; (2) creating a functional expression based on a variational method, and selecting appropriate weighting coefficients; and (3) solving an extremal function of the functional expression by using a splitting Bregman iterative algorithm to obtain a fusion result. The invention overcomes the disadvantages that a high spectral image fusion is not easily to be explanted directly and the retentivity of spectrum is not good in the prior art. The invention achieves a good fusion effect and has a wide application prospect in the technical field of high spectral image fusion.

The invention provides a wheat plant nitrogen content estimation method based on hyperspectral image fusion map characteristics. The method comprises the following steps: acquiring a wheat canopy hyperspectral image and the nitrogen content of a ground wheat plant; firstly, spectral reflectance is extracted, and vegetation indexes, positions and shape features are calculated; secondly, extractinga principal component hyperspectral image and extracting deep features by using a convolutional neural network; a random forest algorithm and a correlation coefficient analysis method are used again to determine optimal features, and a parallel fusion strategy is used to construct new fusion map features for the optimal features; and finally, establishing a support vector regression model based onthe fusion map characteristics to predict the nitrogen content of the wheat plants. The method is high in estimation precision, strong in model generalization and suitable for the whole growth periodof wheat, and is also a method for estimating the nitrogen content of the wheat plant by integrating the vegetation index, the position and shape characteristics and the deep characteristics of the hyperspectral image to construct fusion map characteristics for the first time at present.

The invention provides a wheat leaf layer nitrogen content estimation method based on hyperspectral image fusion map characteristics. The method comprises the following steps: acquiring wheat canopy hyperspectral image data and actually measured wheat leaf layer nitrogen content; firstly, performing image preprocessing, extracting spectral reflectance, calculating vegetation indexes, positions andshape features, and extracting deep features by using a convolutional neural network; secondly, performing feature optimization through correlation coefficient analysis and a random forest algorithm,and constructing new fusion map features by using a parallel fusion strategy; finally, constructing a wheat leaf layer nitrogen content estimation model based on fusion map characteristics by utilizing a particle swarm optimization support vector regression method. The method is high in estimation precision, high in feature robustness and suitable for the whole growth period of wheat, and meanwhile, the method for estimating the nitrogen content of the wheat leaf layer by integrating the vegetation index, the position and shape features and the deep features of the hyperspectral image to construct fusion map features is proposed for the first time at present.

The invention relates to a method and a device for classifying remote images by integrating edge information and a support vector machine. The method includes the steps: performing pixel-wise support vector machine classification for the remote images which are subjected to preprocessing and characteristic extraction; performing edge detection for the remote images which are subjected to preprocessing and characteristic extraction so as to obtain a discontinuous one-pixel wide edge map; performing edge connection for the discontinuous one-pixel wide edge map subjected to noise edge removal so as to obtain a closed edge map; and integrating the closed edge map to a map subjected to pixel-wise support vector machine classification so as to obtain a classified result map integrating the edge information. By the aid of the method and the device, non-robustness of existing partitional-clustering-segmentation-based classification of the hyperspectral images by integrating space and spectral information is overcome, and the problem of dimensional proportion selection of common fixed-window-size-based methods such as morphology filtering by integrating the space and spectral information is solved.

The invention discloses a panchromatic/multispectral remote sensing image and hyperspectral remote sensing image fusion method, relates to the field of remote sensing image processing, and aims to solve the problem that the existing panchromatic/multispectral-hyperspectral image fusion result cannot reasonably balance space and spectral resolution. The method comprises the following specific steps: step 1, extracting an end member matrix of a hyperspectral remote sensing image; step 2, establishing a variational model by using the panchromatic/multispectral remote sensing image, the hyperspectral remote sensing image and the end member matrix; and step 3, carrying out iterative calculation on the variation model by using an alternating direction multiplier method to obtain a fused image of the panchromatic/multispectral remote sensing image and the hyperspectral remote sensing image.

The invention provides a hyperspectral and panchromatic image fusion method based on a cascade network, and belongs to the technical field of image processing, and the method comprises the following steps: integrating a detail injection method and a spectral unmixing model into a cascade convolutional neural network, constructing an end-to-end network model, estimating the abundance of the pixels of a hyperspectral image, and reconstructing a hyperspectral image with high spatial resolution. According to the method, the panchromatic image is introduced and fused with the hyperspectral image, so that the spatial resolution of the hyperspectral image can be improved, and meanwhile, the spectral distortion phenomenon of the fused image is reduced. By establishing a hyperspectral image fusion model based on the cascade convolutional network, a detail injection method and a spectral unmixing model are integrated into an end-to-end network structure, and the space and spectral information of an input multi-source image can be effectively extracted to generate the abundance value of each end member of a high-resolution image.

The invention discloses a hyperspectral image super-resolution reconstruction method based on sparse representation and image fusion. The method comprises the following steps: completing super-resolution reconstruction of a single picture by adopting an approximated Heaviside function sparse representation method, and then further fusing a multispectral image with a hyperspectral image obtained inthe previous step by adopting a local mixed image fusion method to enhance the resolution of the image in consideration of the fact that enough prior information cannot be obtained by only using a low-resolution image.

The invention discloses a target contour extraction method based on fusion of a panchromatic image and a hyperspectral image, and the method comprises the steps: carrying out the sampling of the hyperspectral image to the same spatial resolution of the panchromatic image through interpolation, and obtaining a sampling image on the hyperspectral image; Weighting visible light and a near-infrared band of the sampling image on the hyperspectral image to obtain a low-resolution panchromatic image; generating a panchromatic and hyperspectral fusion image from the sampling image, the low-resolutionpanchromatic image and the panchromatic image on the hyperspectral image; calculating the class density and the distinction degree of each pixel of the panchromatic and hyper-spectral fusion image andthe product of the class density and the distinction degree of each pixel to obtain a central likelihood, and determining a central pixel; And dividing all pixels in the panchromatic and hyper-spectral fusion image into two categories according to the characteristic distances between the pixels and the central pixel, thereby obtaining the contour of the target. According to the invention, throughfusion of the panchromatic image and the hyperspectral image, the spatial resolution of the hyperspectral image is improved.

The invention discloses a multi-hyperspectral image fusion method guided by low-rank prior and spatial spectrum information, and provides a brand new multi-layer multi-branch fusion network SSLRNet combining spatial spectrum guidance and low-rank prior, the network firstly constructs a multi-layer multi-branch fusion sub-network (MLMB), and aims to extract features from a plurality of branches, and then the multi-layer multi-branch fusion sub-network SSLRNet constructs a multi-layer multi-branch fusion sub-network SSLRNet; and multi-layer feature fusion is carried out to reconstruct a preliminary fusion image. And then, constructing a fusion image spatial spectrum correction sub-network based on spatial spectrum guidance, and performing spatial spectrum guidance on a preliminary fusion image generated by the MLMB through a multispectral image waveband superposition summation image and a hyperspectral image waveband average value image, thereby reducing spatial spectrum distortion. And finally, constructing a fusion image low-rank prior constraint sub-network based on a low-rank neural network, combining the sub-network with a deep learning network, and performing low-rank decomposition by utilizing the characteristics of the network, so that a fusion result better meets a real application requirement. According to the invention, the fusion precision of the network is improved, and real application requirements are better met.

The invention relates to an identification method of strokes in calligraphy and paintings. The identification method comprises the following steps: step 1, hyperspectral image formation; step 2, hyperspectral image preprocessing; step 3, hyperspectral image fusion and transformation; step 4, extraction and measurement of stroke features; and step 5, discrimination of similarity of the stroke features. The identification method mainly solves the problem that forgery is difficult to identify when stroke analysis and identification are performed by using human eyes or visible images at present; and stroke information is acquired, modeled and analyzed by the hyperspectral image formation technique and an atlas analysis technique, and space and spectra rather than single spatial information aresubjected to identification simultaneously, thereby finishing objective identification of the strokes in calligraphy and paintings in a more efficient, comprehensive and accurate manner, and better assisting identification of the integral calligraphy and painting works. The identification method is mainly applied to identification of strokes and pencraft in calligraphy, Chinese ink paintings, Chinese color paintings and foreign oil paintings.

The invention discloses a hyperspectral image fusion method combining spectral unmixing prior and learnable degradation constraint, and the method comprises the steps: applying an unmixing model to a hyperspectral image, and building a degradation relation between high-resolution abundance and low-resolution abundance; the low-resolution abundance and the high-resolution abundance are optimized in combination with a hyperspectral image unmixing process; constructing a spatial degradation network and a spectral degradation network to learn a degradation relationship between images, wherein the spatial degradation network can adapt to input of abundance and hyperspectral images at the same time; constructing a residual block to optimize an intermediate image, and learning finer features; the network loss function is an L1 norm. The method has excellent performance when being applied to full-color image and hyperspectral image fusion or multispectral image and hyperspectral image fusion.

The invention discloses a hyperspectral image eigen decomposition method based on digital surface model assistance, belongs to the technical field of hyperspectral image fusion processing, and aims to solve the problem of low precision of existing hyperspectral image eigendecomposition. The method comprises the following steps: inputting a hyperspectral image and digital surface model data, and calculating to obtain geometric components; calculating to obtain a local prior matrix; calculating to obtain a non-local prior matrix; and according to the geometric component, the local priori matrix and the non-local priori matrix, carrying out eigen decomposition, and outputting high spectral reflectivity and environment illumination. The invention is used for intrinsic decomposition of the hyperspectral image.

The invention provides a multi-modal data image fusion classification method based on a graph convolutional network, and mainly solves the problem that the existing hyperspectral image classification precision is low. According to the implementation scheme, a hyperspectral image data set is obtained; performing multi-source data fusion on the original HSI and LiDAR images in a GS fusion mode to obtain a hyperspectral image GS fusion image; respectively extracting invariant attribute features of the HSI image and the LiDAR image, and obtaining an invariant attribute feature fusion image through a feature fusion mode based on a weighted fusion image; inputting the hyperspectral image GS fusion image and the invariant attribute feature fusion result into miniGCN and 2DCNN branches to further extract spectral features and spatial features; carrying out feature fusion on the extracted spectral features and spatial features through a multiplicative fusion strategy; and classifying multiplicative feature fusion results through a classification network. The method reduces the loss of feature information, improves the classification performance, and can be used for hyperspectral image classification.