356 results about "Image noise reduction" patented technology

Certain embodiments of the present invention provide a system and method for reducing image noise with the use of a minimal error spatiotemporal recursive filter. An input image is filtered both temporally and spatially, producing a temporal output and a spatial output. Both the temporal output and the spatial output are correlated with the input image to produce a temporal correlation output and a spatial correlation output. The temporal correlation output and the spatial correlation output are mixed to generate a selecting signal. The selecting signal directly or indirectly determines the composition. of an output image. The selecting signal may select a portion of the temporal output and a portion of the spatial output to compose an output image. Alternatively, the selecting signal may select either the temporal output or the spatial output to compose an output image.

A method and apparatus for processing imager pixel signals to reduce noise. The processing includes receiving a target pixel signal, receiving at least one neighboring pixel signal, formulating a dynamic noise signal based at least in part on a value of the target pixel signal, and controlling a noise reduction operation using the dynamic noise signal.

The invention relates to an image noise reduction method utilizing a depth full-convolution coding-decoding network. According to the method, a neural network model is employed to carry out image noise reduction, input of the model is an image with noise, and output of the model is the image after noise reduction. The method is characterized in that the model employs a symmetric coding-decoding network structure and comprises N convolution layers and N de-convolution layers which are sequentially connected, an ith convolution layer and an (N+1-i)th de-convolution layer are in symmetric relationship, i=1, 2, and so on, and N, the convolution layer is taken as a characteristic extractor and is used for coding main content of the image and carrying out noise reduction, and the de-convolution layer is used for decoding abstract content of the image and recovering detailed content of the image. Compared with the prior art, noise reduction is carried out while the main content of the image is coded, moreover, the de-convolution layer is used for decoding the abstract content of the image and recovering the detailed content of the image, and the detailed content of the image can be kept to the maximum degree.

The invention provides a small infrared aerial target detection method based on non-downsampling contourlet transformation. The method includes the following steps of 1, non-downsampling contourlet transformation, wherein non-downsampling contourlet transformation first-level decomposition is performed on a small infrared target image, and a band pass sub-band is discomposed into four-direction high-frequency sub-bands; 2, background suppression, wherein low-frequency influences are removed, and thresholding processing is performed on a high-frequency portion; 3, coefficient mapping, wherein coefficients left by the four-direction high-frequency sub-bands are mapped to a gray level space in a linear mode; 4, high-frequency image segmentation, wherein four-direction high-frequency sub-band images are segmented into binaryzation images; 5, binary high-frequency image noise reduction, wherein small bright noise points in the binary high-frequency images are eliminated; 6, detection of related small targets in dimension, wherein the four-direction high-frequency sub-band images get along with each other to obtain a small target single-frame detection result; 7, small target sequence detection, wherein comprehensive vote is performed on multi-frame images to intercept and capture small targets. According to the method, the problem of small aerial target detection under the complicated infrared background is solved.

The invention discloses a medical image noise reduction method based on structure clustering and sparse dictionary learning, which overcomes the problems that in the prior art, the dictionary learning method neglects the structure similarity of image blocks, so that detailed information of images is lost, and the homogeneous region of images is not smooth. The medical image noise reduction method based on structure clustering and sparse dictionary learning, which is provided by the invention, comprises the following implementation steps: (1) inputting noise images; (2) pre-filtering; (3) clustering structures; (4) extracting various training sample sets; (5) learning sparse dictionaries in classification; (6) acquiring the final training dictionary; (7) acquiring various noise reduction estimation values; (8) acquiring images after noise suppression; (9) outputting images after noise suppression. The medical image noise reduction method based on structure clustering and sparse dictionary learning, which is provided by the invention, has the advantages of effectively removed noise, improved vision effect of images, effectively kept edges and patterns, and homogenized regional smoothness, and has the capability of being applied to medical computed tomography and noise reduction treatment of nuclear magnetic resonance images.

The embodiment of the invention discloses an image noise reduction method and apparatus, a storage medium and a terminal. The method comprises the following steps: determining brightness information of a skin color area in a target image; determining a target subarea with brightness lower than a preset brightness threshold in the skin color area, and determining target noise reduction intensity according to the brightness of the target subarea and the brightness information; and performing noise reduction processing on the target subarea based on the target noise reduction intensity to obtaina target image after noise reduction processing. By adoption of the above technical scheme, the target subarea to be performed with the noise reduction processing is determined based on the brightness, the noise reduction intensity is determined according to the brightness of each pixel point in the target subarea and the brightness information of the skin color area, and the noise reduction processing is performed on the corresponding pixel point based on the noise reduction intensity to achieve a local noise reduction effect of the skin color area based on the brightness, so that the noisepoint distribution of the skin color area is more uniform.

The invention discloses a method for removing image noise based on kernel regression total variation, and belongs to the field of computer vision. The method comprises the steps: (1), enabling a kernel regression method to be extended in a bounded total variation space, and constructing a regularization item comprising noise image local priori knowledge; (2), constructing a regularization item of non-local priori knowledge through employing non-local similar information in an image, and building an image noise reduction model; (3), carrying out the quick solving of the constructed model through employing a split Bregman method, and obtaining a preliminary estimated image; (4), carrying out the residual error iteration optimization of the preliminary estimated image, and achieving the optimal image noise removing effect. The method keeps the detail and texture information of the image well while maintaining an image structure.

The invention discloses an image noise reduction method, overcoming the problem of fixed bias between a wavelet coefficient obtained by an image noise reduction method in the existing image processing technology and a wavelet coefficient of an original image. The image noise reduction method comprises the following steps of carrying out multilevel wavelet decomposition on to-be-processed noise images, thereby acquiring the corresponding multilevel wavelet coefficients; according to the wavelet coefficients at all levels and the corresponding level numbers of the wavelet coefficients, determining the corresponding noise thresholds of the wavelet coefficients at all levels; carrying out noise reduction on the multilevel wavelet coefficients by adopting a corresponding wavelet threshold noise reduction function of multiple noise thresholds based on the multilevel wavelet coefficients; and reconstructing the original images corresponding to the noise images by adopting the noise-reduced multilevel wavelet coefficients. The constant error among the wavelet coefficients obtained by the image noise reduction method and the wavelet coefficients of the original images is relatively small so that pseudo-Gibbs artifacts can be avoided, detailed information of the images are well retained, and the calculated quantity is relatively low. Thus, the image noise reduction method can be widely applied to the wireless broadcasting field.

The present invention discloses an enhancement processing method and apparatus for a traffic video image. According to the method and the apparatus, the noise intensity is estimated and an image noise reduction process is guided according to continuous video image inter-frame information, a constraint weight of a neighborhood pixel region to a center pixel unit is introduced according to extracted local feature effective data, and a grey-scale mapping relationship function of the pixel unit is adaptively constructed, so that image noise suppression is more accurate, image local detail enhancement is more obvious, and further the effectiveness of a video image enhancement method in a complex scenario, especially a low-illumination environment is improved.

The invention provides a image quality improving apparatus which comprises: an exposure acquiring module for acquiring the exposure of image; a saturation regulating module for increasing or decreasing the image saturation according to the exposure; acutance regulating module for increasing, decreasing or eliminating the image or doing the anti-acutance according to the exposure; a contrast regulating module, for increasing or decreasing the image contrast according to the exposure; a color enhancement regulating module, for increasing or decreasing the image color intensity, or eliminating the color enhancing function according to the exposure; and a noise reduction module for opening, closing or reducing the image noise reduction function. Processed by the apparatus, the image becomes color saturated, verisimilar, clear and low noised.

The invention discloses a method for automatically detecting the adhesion defect of a label of a flat wine bottle. According to the method, a lighting image acquisition system, an image processing system and an algorithm processing system are adopted, wherein the lighting image acquisition system is used for acquiring an image of a labeled wine bottle; the image processing system is used for preprocessing the obtained image of the wine bottle through image contrast stretching conversion, image noise reduction processing, image threshold processing and image filling processing in sequence; the algorithm processing system is used for positioning a body of the wine bottle and a set point in the image of the wine bottle, and comparing a position relation between a special point during correct adhesion of a label and the label with a position relation between a special point in the image needing to be detected and the label so as to judge whether the label has the adhesion defect. The method can be used for detecting the adhesion defect of the label of the flat wine bottle, which cannot be detected through a common label detection method and an ordinary visual detection method.

The invention discloses an image recognition technology-based forest fire automatic monitoring and recognition system and method, and belongs to the field of image processing. Based on the characteristics of wide application range and high economy of video monitoring image processing and mode recognition technologies, the image recognition technology-based forest fire automatic monitoring and recognition system and method are applied to the field of forest fire monitoring and early warning. The image recognition technology-based forest fire automatic monitoring and recognition method comprisesthe following steps: performing preprocessing of image noise reduction and data compression by using median filtering and gray processing methods; detecting a moving target in the monitoring image byusing a Gaussian mixture background model and a differential detection technology; extracting multi-dimensional features such as color moment, energy, entropy, inverse difference moment and circularity from the detected target image, and fusing the multi-dimensional features into a normalized feature vector; and finally, using a binary classifier of a support vector machine for identification andjudgment. According to the image recognition technology-based forest fire automatic monitoring and recognition system and method, a moving target appearing in a complex background can be well detected, and characteristic parameters are extracted for identification, and automatic fire identification is further realized, and the image recognition technology-based forest fire automatic monitoring and recognition system and method have the advantage of high monitoring accuracy.

The invention provides an anti-printing shot image digital watermarking method based on image noise reduction. A generative adversarial network is adopted to construct an image noise reduction layer capable of resisting possible noise attacks of an image in the printing and shooting process, simultaneous attacks of various kinds of noise can be resisted, a better noise reduction effect is achieved, and it is guaranteed that the image has high fidelity; an image noise reduction layer is added into the whole watermark embedding and extracting training framework; the image noise reduction layer achieves the function of resisting noise attacks, and ensures robustness to a certain extent, so that the watermark codec can focus on improving the visual effect after watermark embedding, the watermark detection accuracy and the embedding capacity, and the balance of three indexes of robustness, the visual effect and the embedding capacity is realized.

The present invention discloses a cement notch groove pavement image noise reduction enhancement and crack feature extraction method aiming at the problems that the pavement contrast is too low causedby external factors and the payment spots and notch groove autointerference caused by pavement materials. The method comprises the following steps of: employing an improved local adaptive contrast enhancement algorithm to enhance image contrast after graying processing of an original cement pavement image; employing translation invariance Shearlet transform denoising algorithm of an improved P-Mmodel to remove speckle noise caused by pavement materials; employing a cement notch groove pavement image smoothing model established based on an unidirectional total variation UTV model to the imageafter denoising to remove a pavement notch groove influencing feature extraction; and combining a connected domain mark method, a projection method and a rectangular frame method to extract a crack type determination method and a crack feature calculation method to achieve digital description of crack features. The cement notch groove pavement image noise reduction enhancement and crack feature extraction method is systematic and comprehensive, small in calculated amount and easy to apply.

The invention discloses a method for inhibiting an image noise based on an optimum primary color signal. A digital camera executes this noise inhibition method to inhibit the noise in the digital images mainly comprising: taking out the primary color signal in the digital image, taking out the high and low frequency signals of the optimum primary color signal through a high-pass filter and a low-pass filter, and the high and low frequency signals of the worst primary color signal, then mixing the high frequency signal of the optimum primary color signal and the worst primary color signal, and stacking the mixed signals to the low frequency signal of the worst primary color signal to obtain the digital image signal after the noise inhibition.

The invention discloses a method and device for noise reduction of an image. The method for noise reduction of the image comprises the steps that the image I0 is input, noise reduction based on weighted average is conducted on the image I0, an initial smooth image I1 and a weighted image W1 are obtained, and the weighted image W1 is composed of smooth weights W1 (x, y) of all pixels in the smooth image I1; a target weight Wt is selected based on the weighted image W1; the smooth weight W1 (x, y) of each pixel and the target weight Wt are compared, the pixels with the smooth weights W1 larger than the target weight Wt are corrected reversely, and the pixels with the smooth weights W1 smaller than the target weight Wt are corrected forwardly. According to the method and device for noise reduction of the image, in combination with the smooth weights of all the pixels of the image and the target weight obtained based on the smooth weights, a region which is excessively smooth and a region where noise reduction is incomplete in the image are selected, and the two regions are processed respectively; as a result, noise is reduced effectively, the problem that the smoothness degrees of parts of the image are not uniform in the noise reduction process of the image is solved, and noise reduction of the image is more uniform.

The invention discloses a gray level image noise reduction method based on hole convolution and an automatic coding and decoding neural network, which comprises the following steps of taking a clear image and a noise image corresponding to the clear image as a training sample to construct a training set; constructing an image noise reduction model, wherein the image noise reduction model comprisesan image feature coding unit for performing feature coding on a noise image, the image feature decoding unit is used for decoding the coded feature map output by the image feature coding unit, the image feature coding unit comprises a feature extraction module and 10 coding modules, and the image decoding unit comprises 10 decoding modules and an image restoration module; utilizing the training set to train the constructed image noise reduction model to obtain a trained image noise reduction model; during application, inputting a noise image into the trained image noise reduction model, and outputting a noise reduction image through calculation. The grayscale image noise reduction method can quickly remove the noise of the image and improve the visual effect of the image.

The invention discloses a low-illumination-image noise reduction method. The method includes: acquiring a low-illumination image; segmenting the low-illumination image into different texture regions according to differences of image texture information, and dividing the texture regions into foreground image regions and background image regions; adopting different processing strategies to respectively carry out noise reduction on the foreground image regions and the background image regions to obtain processed foreground image regions and background image regions; and synthesizing the processedforeground image regions and background image regions to obtain a noise-reduced image. The invention also discloses a low-illumination-image noise reduction device.

The invention discloses an image noise reduction device and method, and the method comprises the steps: obtaining a plurality of continuous frame images photographed by a terminal, enabling one of the frame images to serve as a sample image, and enabling the other frame images to serve as to-be-compared images; respectively obtaining the brightness values of R, G and B components of all pixels of the sample image, and the brightness values of R, G and B components of all pixels of the to-be-compared images; and obtaining the image after noise reduction according to the mean value of the brightness values of R components of all pixels of the sample image and the to-be-compared images, the mean value of the brightness values of G components of all pixels of the sample image and the to-be-compared images and the mean value of the brightness values of B components of all pixels of the sample image and the to-be-compared images if the difference between the brightness value of the R components of all pixels of the sample image and the to-be-compared images, the difference between the brightness value of the G components of all pixels of the sample image and the to-be-compared images and the difference between the brightness value of the B components of all pixels of the sample image and the to-be-compared images are within a first preset range. The method and device solve problems of large calculation burden and poor effect of noise reduction in a noise reduction process of the image, and also solve a problem that the details of the edge of the image are liable to be lost.

The invention discloses an efficient image restoration method based on an alternating direction method. The efficient image restoration method based on the alternating direction method includes: translating image restoration problems into low rank matrix complement problems; further dividing each low rank matrix into a plurality of partitioned matrixes by decomposing an original image into three low rank matrixes according to R/G/B channels; complementing each partitioned matrix; combining all the partitioned matrixes after being complemented so as to obtain corresponding complement matrixes; further obtaining a repaired image according an combination of the complement matrixes. The efficient image restoration method based on the alternating direction method facilitates parallel processing through partitioning, complements each partitioned matrix by using the alternating direction method, greatly improves image restoration speed, can not damage overall relevance of the original matrixes due to the fact that the plurality of the partitioned matrixes are obtained by dividing the three low rank matrixes according to relative independence of the three low rank matrixes, and facilitates improvement of restoration accuracy. The efficient image restoration method based on the alternating direction method not only can be used for image restoration, but also can be used in fields of object removal, image noise reduction and the like.

A method for image noise reduction method, device, computer readable storage medium and electronic device are provided.The method comprises the following steps of: acquiring multiple frames of original images, screening the multiple frames of original images according to the definition of the multiple frames of original images, retaining the clear target images, determining the noise suppression degree according to the number of frames of the screened original images, and finally synthesizing and denoising the retained target images according to the noise suppression degree. As a result, the noise suppression degree is determined according to the number of frames screened out from a plurality of original images, and the reserved images are synthesized and denoised according to the determined noise suppression degree, which not only effectively reduces the noise of the images and retains the information of the images to the maximum extent, but also improves the quality of the photographed images and improves the user experience.

The embodiment of the invention discloses an image noise reduction method and device, a storage medium and electronic equipment, and the method comprises the steps: obtaining a plurality of to-be-processed images of the same photographing scene; determining a first image as a reference image from the multiple frames of to-be-processed images, and taking other images except the first image as second images; respectively segmenting the first image and the second image into a plurality of regions; calculating a first homography matrix of the second image in each area relative to the first image;and performing noise reduction fusion processing on the second image and the first image on each region based on the corresponding first homography matrix to obtain a target image. According to the scheme, a mode of performing regional processing on the images is adopted, so that the ghosting phenomenon in the synthesized image is eliminated while multi-frame synthesis noise reduction is realized.

The invention discloses a medical image processing method and system. The method comprises the steps that a medical image to be processed is acquired, and the interested region of the medical image is extracted; filtering and parameter optimization are carried out on the interested region to acquire an optimal filtering parameter; the optimal filtering parameter is used to filter regions outside the interested region of the medical image; and the filtered medical image is output. According to the technical scheme, the optimal filtering parameter is acquired from the interested region; the optimal filtering parameter is applied to global image filtering; and the method has the advantages of stable filtering effect, consistent whole image noise reduction and fast processing speed. In addition, the invention further provides a medical image processing system.