399 results about "Gaussian blur" patented technology

System and method of generating feature descriptors for image identification. Input image is Gaussian-blurred at different scales. A difference of Gaussian space is obtained from differences of adjacent Gaussian-blurred images. Key points are identified in the difference-of-Gaussian space. For each key point, primary sampling points are defined with three dimensional relative positions from key point and reaching into planes of different scales. Secondary sampling points are identified for each primary sampling point. Secondary image gradients are obtained between an image at a primary sampling point and images at secondary sampling points corresponding to this primary sampling point. Secondary image gradients form components of primary image gradients at primary sampling points. Primary image gradients are concatenated to obtain a descriptor vector for input image. Descriptor vector thus obtained is scale invariant and requires a number of additions equal to number of primary sampling points multiplied by a number of secondary sampling points.

A method of and an apparatus for determining a depth map utilizing a movable lens and an image sensor are described herein. The depth information is acquired by moving the lens a short distance and acquiring multiple images with different blur quantities. An improved method of simulating gaussian blur and an approximation equation that relates a known gaussian blur quantity to a known pillbox quantity are used in conjunction with a non-linear blur difference equation. The blur quantity difference is able to be calculated and then used to determine the depth map. Many applications are possible using the method and system described herein, such as autofocusing, surveillance, robot / computer vision, autonomous vehicle navigation, multi-dimensional imaging and data compression.

A method of and an apparatus for determining a depth map utilizing a movable lens and an image sensor are described herein. The depth information is acquired by moving the lens a short distance and acquiring multiple images with different blur quantities. An improved method of simulating gaussian blur and an approximation equation that relates a known gaussian blur quantity to a known pillbox quantity are used in conjunction with a non-linear blur difference equation. The blur quantity difference is able to be calculated and then used to determine the depth map. Many applications are possible using the method and system described herein, such as autofocusing, surveillance, robot / computer vision, autonomous vehicle navigation, multi-dimensional imaging and data compression.

The invention provides a blurred photos generation method, blurred photos generation device and a mobile terminal. The blurred photo generation method comprises steps of detecting a contour edge of a shooting subject in a foreground area in a current preview image, obtaining a blurring coefficient for blurring processing on the background area beyond the couture edge in the current preview image, and performing gauss blurring processing on the background area according to the blurring coefficient to generate a blurred photo. The blurred photo generation method can prevent the shooting subject from being blurred or can prevent the background from not being completely blurred and improves blurring accuracy of the photo.

Herein is presented a catadioptric projector by combining a commodity digital projector with additional optical units. By using specially shaped reflectors and / or refractors, a catadioptric projector can offer an unprecedented level of flexibility in aspect ratio, size, and field of view. Also presented, are methods to reduce projection artifacts in catadioptric projectors, such as distortions, scattering, and defocusing. By analysis of projection defocus of reflector and thin refractor based catadioptric projectors, it is shown that defocus blur can be interpreted as spatially-varying Gaussian blurs on an input image. Kernels are measured directly from a light transport matrix, T, and de-convolution is applied to optimize an input image. Practical uses of catadioptric projectors in panoramic and omni-directional projections are also demonstrated.

A digital imaging system and method uses a two-stage sigma filter to de-screen color images. This filter does not assume any a priori knowledge about the screening process using to produce the halftone image. The two-stage sigma filter may therefore be used to convert color halftone images into continuous-tone images irrespective of the screening process used. The two-stage sigma filter may be constructed, or emulated in software, using an O (N) algorithm which performs smoothing and preserves edge information simultaneously in the Red / Green / Blue color space. This system and method outperforms conventional approaches which, for example, use a Gaussian blur, because it satisfies the dual criteria of completely eliminating halftone screens while preserving edge information. When combined with halftone segmentation techniques, a complete document processing algorithm for gray-scale and color documents is created.

The invention relates to a multipoint touch method based on binocular stereoscopic vision, which relates to the technical field of computer vision. The method comprises the following steps: firstly, the azimuth and the projection parameters of a camera relative to a three-dimensional scene are determined through camera calibration, and a binocular camera is utilized to acquire an image pair; secondly, two pictures on the same frame which are acquired on the first step are taken as input, and a scene depth map is calculated by the stereoscopic vision algorithm; thirdly, closest patches are obtained through Gaussian blur and determined thresholds, and central coordinates of the patches are calculated; and fourthly, a plurality of coordinates obtained in the third step are taken as input of a multipoint touch program, and interactive results are returned to a screen. The multipoint touch method based on binocular stereoscopic vision can realize completing the multipoint touch operation in space by a user without touching the screen, and coordinate information of input points acquired is three-dimensional, so that the multipoint touch method can realize more complex interaction such as push and pull of a virtual object and so on.

The present invention proposes a bokeh photograph generation method and device, and a mobile terminal. The bokeh photograph generation method comprises the following steps: determining first depth of field information about a foreground region and second depth of field information about a background region in a current preview image according to preview image data respectively acquired by two rear-facing cameras in a dual-camera device; acquiring a base value of bokeh degree according to the first depth of field information and the second depth of field information; and performing Gaussian blur processing on the background region according to the base value of bokeh degree to generate a bokeh photograph. The bokeh photograph generation method proposed by the present invention can simplify the process of generating a background bokeh photograph when photographing, thereby improving the shooting effiency.

The invention discloses a rapid license-plate positioning method based on definition and luminance evaluation. The rapid license-plate positioning method includes the following steps: 1, carrying out definition evaluation based on noise and sharpness on an input image; 2, carrying out gradient sharpening processing if the definition is insufficient, and carrying out Gaussian fuzzy processing if the definition is excessively high; 3, converting the RGB image into a grey-scale image; 4, carrying out luminance evaluation on the grey-scale image; 5, carrying out illumination unification processing if the luminance is abnormal; 6, extracting the perpendicular edges in the image through a Scharr operator; 7, carrying out local self-adaption threshold processing; 8, filtering the noise through morphological processing, and fusing the perpendicular edges to form a communicated area; 9, carrying out area marking on the communicated area, carrying out screening according to the characteristics of Chinese license plates, and obtaining a license plate area.

The invention provides a natural background video sectional drawing method. The method comprises the steps that 1) the method automatically selects a key frame, specifically the method automatically selects the key frame in a video by calculating inter-frame difference; 2) a key frame Trimap map is generated, specifically a user us allowed to provide useful information through a small amount of manual interaction and a GrabCut segmentation algorithm and a Gaussian blur method are used to semi-automatically generate the Trimap map on the key frame; 3) a middle frame Trimap map is generated, specifically bidirectional optical flow propagation is used to generate the Trimap map of each middle frame, and color, gradient and other information are used to correct errors generated by optical flow propagation; and 4) according to the Trimap maps of all video frames and a Bayesian sectional drawing algorithm with added three-dimensional space-time optimization, the alpha mask map of each frame of the video is calculated, and the alpha mask map and a new background are synthesized to acquire a new synthesized video. According to the invention, the method has the advantages of fast speed and less user interaction, and the finally acquired new synthesized video has the advantages of good time coherence, high reality and the like.

The invention discloses a method for improving image definition in foggy days based on an imaging model. The method for improving the image definition in foggy days based on the imaging model includes the following steps that firstly, an RGB image of original misty days is converted into a YUV color space and three-grade wavelet decomposition is carried out on the luminance Y component; a misty day imaging model is combined on a low-frequency sub-band of a wavelet zone, Gaussian Blur is used for estimating and removing scattered light of media, attenuation factors are adjusted with a method based on local complexity and self-adaption enhancement is performed on an attenuation low-frequency sub-image; a non-linear conversion enhancement method is utilized on a high-frequency sub-band to further enhance high-frequency information; finally, a sharpened colored image is obtained through wavelet inverse conversion and conversion of the color space. According to the method for improving the image definition in foggy days based on the imaging model, the problem of blurring caused by atmospheric scattering can be rapidly and effectively solved, the definition and the contrast ratio of the foggy day image are effectively improved and high real-time performance is achieved.

The invention discloses an electronic image stabilizing method and an electronic image stabilizing system aiming at moving object detection under camera shake. The method comprises the following steps: carrying out guassian blur on the original image, lowering the resolution ratio and determining three layers of images; determining a first translational motion vector corresponding to an image with the lowest resolution ratio positioned at the highest layer based on a blocking gray projection method; compensating an interlayer image based on the acquired first translational motion vector; determining a second translational motion vector corresponding to the image positioned in the interlayer based on a block match algorithm; compensating the original image with the highest resolution ratiopositioned at the lowest layer based on the first translational motion vector and the second translational motion vector; and smoothing and compensating the motion by utilizing a feature matching method. The method and system provided by the invention have good robustness to the dynamic background, the noise, and the consistency of motion, rotation and background of large-sized images, have superiority on the image quality, the visual effect and the calculation speed, and satisfy the requirements of the moving object detection under the camera shake after the image is stabilized.

The provided is an image processing method, which includes: converting a to-be-processed image into a gray image; performing Guass fuzzy processing having a predefined fuzzy radius on the gray image, and obtaining a reference image; and according to gray values of reference pixels of the reference image, decreasing a pixel value of a pixel of the to-be-processed image corresponding to a reference pixel whose gray value is larger than a gray threshold, increasing a pixel value of a pixel of the to-be-processed image corresponding to a reference pixel whose gray value is smaller than the gray threshold, and obtaining an adjusted image.

The invention relates to a foreground segmentation method based on significance detection, which comprises the following steps of: (1) inputting a color image in an RGB (Red, Green, Blue) format; (2) subjecting the color image in the RGB format to median filtering; (2) extracting color characteristics, brightness characteristics and direction characteristics to obtain color characteristic patterns, brightness characteristic patterns, and direction characteristic patterns, which form a characteristic pattern set containing eight characteristic patterns; (4) subjecting 8 patterns in the characteristic pattern set to multi-scale sampling; (5) subjecting each characteristic pattern to Gaussian filtering, and calculating the mean value of each characteristic pattern; (6) calculating Euclidean distance between the Gaussian blurred image and the mean value of each characteristic pattern; (7) obtaining color salient patterns, brightness salient patterns and direction salient patterns; (8) blending the above salient patterns to obtain a comprehensive salient pattern; and (9 ) subjecting the comprehensive salient pattern to the foreground segmentation by means of K-means clustering. The foreground segmentation method can effectively inhibit the interference of noise and background on a target and has pretty good real time performance.

The invention discloses a device and a method for virtually processing a depth image. The device comprises an obtaining module used for obtaining an aperture value, a focal length, a focusing distance and a permissible confusion circle diameter; a calculating module used for calculating a field depth range according to the obtained aperture value, focal length, focusing distance and permissible confusion circle diameter; a determining module used for obtaining object distances to a focus and determining virtual degrees corresponding to the different object distances according to the obtained object distances to the focus, aperture value and focusing distance; a correction module used for correcting radiuses corresponding to Gaussian blur according to the determined virtual degrees corresponding to the different object distances; and a fuzzy processing module which is used for carrying out fuzzy processing on pixels of the depth image requiring virtual processing and out of the field depth range according to the corrected radiuses corresponding to the Gaussian blur and is used for shooting. According to the technical scheme provided by the invention, the shooting effect is improved without changing a physical hardware structure of an existing mobile terminal.

The invention provides a facial image layer decomposition method based on an improved guide filter, and the method comprises the steps: carries out the improved guide filtering of a facial image brightness channel, obtaining a human face structure information layer, further carrying out differential operation, and obtaining a texture detail layer. According to the invention, the method carries out the regularization factor adaptive adjustment optimization of a guide filter and Gaussian blurring and corrosion processing of a parameter matrix image, and obtains different filtering effects of a facial image in different regions. The method is used for a digital makeup system, obtains an excellent layer decomposition effect, and remarkably reduces the time complexity of layer decomposition.

The invention discloses a measurement method for image Gaussian Blur. The method is that the edge width of a Gaussian Blur image acts as a parameter for judging the image blurring degree. The implementation includes the steps of acquiring a grayscale of an image; acquiring a gradient magnitude diagram and a gradient direction angle of the grayscale; constructing a multi-scale and multi-direction Gaussian-Laplace filter; acquiring an energy diagram and a scale diagram correspondingly by carrying out convolution on the gradient magnitude diagram and the multi-scale and multi-angle Gaussian-Laplace filter; determining the position of an edge point through acquiring the position of a maximum value of each pixel on the energy diagram in the gradient direction; and finally acquiring the edge width of the original grayscale through calculation, thereby realizing measurement of the image Gaussian Blur. According to the method disclosed by the embodiment of the invention, the ratio between the increasing rate of the edge width acquired according to the method and the increasing rate of the blurring degree maintains certain stability, thereby avoiding the circumstance that an error occurs in a traditional evaluation method.

The invention provides a video content replacement method and system based on an adversarial generative network. The method comprises the following steps: extracting a source target image in a video frame; performing semantic segmentation on the source target image; performing data enhancement on the source target image by using the image transformation operation; inputting the semantic segmentation images paired by the enhanced data set into a network training generative adversarial model; accurately detecting an ROI region of the source target image and carrying out cutting and straightening; taking the ROI image as a generation model input to obtain a generated target image and a synthetic mask; controlling edge smoothing and deblurring of the generated target image by using mask control after Gaussian blur; adjusting the hue of the generated target image to be consistent with that of the frame source target image by a histogram matching strategy; generating video frame anti-jitterprocessing; and fusing the source image and generating a target image. Compared with the prior art, the method and system has the characteristics and advantages of easiness in operation, low cost, good effect, short consumed time and the like.

An improved sharpening technique sharpens an image using diffusion rather than a Gaussian blur in the smearing phase. After an appropriate radius or number of passes is determined for the pixels of an image, a diffusive process is applied to the image in the smearing phase to generate a blurred image. The values of each pixel in the original image are differenced with the values of each corresponding pixel in the blurred image. The difference may be weighted by input from a user. The resulting value is added to the original image and a sharpened version of the original image is created.

The invention provides a self-adaptive image scaling method based on bicubic interpolation. The self-adaptive image method based on the bicubic interpolation comprises the following steps: conducting a Gaussian blur operation of a source image so that detail information which can not be displayed under the condition of low resolution is eliminated and distortion of a scaled image is avoided; finding the corresponding pixel location of a target image pixel point on the source image, self-adaptively selecting m sampling pixel points which are corresponding to the periphery of pixel points, obtaining a weighted value which is corresponding to each sampling pixel point according to a bicubic interpolation function, and weighting and summing and to get a pixel value of the scaled image according to the weighted values of the m sampling pixel points and pixel values of the m sampling pixel values; and conducting a sharpening operation of the scaled image so that the edges of an image are clear and a target image is obtained.

The invention discloses a camera wide dynamic image enhancement method and a device. Pixel mean values of a bright pixel area and a dark pixel area of images in original image data are respectively and statistically figured out, brightness increase factors are obtained by means of different value ranges of the pixel mean values, the brightness increase factors are used as parameters and an index curve is used so that original images are mapped once to obtain stretched images, the stretched images are subjected to gaussian blur processing to obtain low frequency image spectrum of the stretched images, and a specific value of the low frequency image spectrum to the stretched images is used as a parameter to perform contrast ratio improvement for the stretched images so as to finally obtain wide dynamic images.

The invention relates to the technical field of camera focusing, in particular to an automatic focusing method and device for a camera based on a defocus estimation improved hill climbing method. Thedefocus state of the camera can be regarded as a Gaussian blur state of an image. When the camera is in a quasi-focal state, the original image of a current frame image deviates greatly from a definition evaluation value of the Gaussian blurred image; when the camera is in the defocus state, the original image of the current frame image deviates little from the definition evaluation value of the Gaussian blurred image, and if the defocus depth is larger, the deviation of the definition evaluation value is smaller. Thus, the defocus depth of the current frame image can be expressed by calculating the deviation between the original image of the current frame image and the Gaussian blurred image of the current frame image, and the focusing step of the camera is adaptively adjusted according to the defocus depth, so that the quasi-focus position can be quickly arrived at on the premise that the focusing precision is ensured, and the problem that the speed and the precision of the camera are difficult to balance during focusing is solved.

The invention discloses an objective evaluation method for the quality of stereo images. The method comprises the following steps of: firstly, respectively extracting a characteristic vector of Gaussian Blur distortion level reflecting the image quality, a characteristic vector of white noise distortion level reflecting the image quality, a characteristic vector of JPEG (Joint Photographic Experts Group) distortion level reflecting the image quality and a characteristic vector of JPEG2000 distortion level reflecting the image quality of the stereo images, so as to obtain a characteristic vector of the stereo images; and testing each distorted stereo image of the same distortion type by a support vector regression training model to get an objective quality evaluation predictive value of each distorted stereo image. The method has the advantages that, in the method, based on the influence of different distortion types on the distorted stereo images, the characteristic information of different distortion types is extracted to form the characteristic vector of the stereo images, so that the obtained characteristic information of the stereo images has strong stability and can reflect the quality variation status of the stereo images better, and the relevancy of the objective evaluation result and the subjective perception is improved.

The invention relates to an image enhancement method achieved by intelligently increasing the field depth. The method includes the steps of obtaining a practical body area by marking an original image to obtain a marked area and by conducting outward judgment on the boundary of the marked area, then, generating a masking layer on the marked area or the body area, and finally conducting mixed calculation on the original image and a Gaussian blur image according to the generated masking layer to obtain a result image with the intelligently-increased field depth. Therefore, intelligent field depth increase can be rapidly and accurately conducted on the image, and the effect can be more natural.

The invention discloses an image depth extraction method, which comprises the following steps of: (1) performing Gaussian fuzzy processing on an original image to be processed to obtain N fuzzy images, wherein N is more than or equal to 2; (2) detecting the edges of the original image and the N fuzzy images to obtain an edge image of each image; (3) calculating the corresponding fuzzy parameter estimation value of each pixel point under each Gaussian filtering parameter on the edges based on the edge images; (4) analyzing N fuzzy parameter estimation values of each pixel point on the edges by using a statistical method to obtain an optimal fuzzy parameter estimation value of each pixel point on the edges; (5) calculating the depth value of each pixel point on the edges in the images according to the optimal fuzzy parameter estimation value of each pixel point on the edges to obtain a sparse depth map; and (6) performing interpolation processing on the sparse depth map to obtain a dense depth map. Compared with a method in the prior art, the method has the advantages that high-accuracy fuzzy parameters can be obtained, so that a depth value obtained through subsequent calculation has high accuracy.

The invention belongs to the field of digital image processing and discloses a method for automatically removing image black eyes. The method is characterized by comprising the following steps: first drawing the contour regions of a left eye bag and a right eye bag and marking as eye bag contour pictures; then, obtaining a face region picture; extracting the brightness of each pixel of the face region, leading into a result and marking as an image B; sequentially carrying out average filtering, high contrast reservation and brightening algorithms on the image B to obtain an image C; sequentially carrying out average filtering and histogram statistics on the image B, then calculating the image B with the eye bag contour pictures and the image C to obtain an image D, and carrying out Gaussian blurring on the image D; carrying out Gaussian blurring on the face region to obtain a face Gaussian picture; finally calculating the image D, the face region and the face Gaussian picture to obtain a result picture. The technical scheme realizes the function of automatically removing black eyes, reduces user operation difficulty and threshold to the lowest, can be popularized in various portable digital devices, saves man-machine interaction devices such as a mouse and a keyboard, is low in hardware cost and has the characteristics of convenience and rapidness.

The invention discloses a crop growth situation monitoring Internet of Things system based on visual inspection. The crop growth situation monitoring Internet of Things system comprises a plant height monitoring subsystem, a lodging monitoring subsystem and a leaf area index monitoring subsystem. The plant height monitoring subsystem is used for obtaining the relative distance between a crop and a height reference mark by setting the preset height reference mark and then obtaining the actual height of the crop according to the mapping relationship between the relative height and the actual crop height. The lodging monitoring subsystem is used for carrying out ashing processing, Gaussian Blur processing and binarization processing on an original image in sequence and then calculating a lodging image and further calculating the lodging rate according to the result of binarization processing. The improvement on precision and efficiency is facilitated through an automatic image analysis manner. The leaf area index monitoring subsystem is used for calculating the comprehensive coverage degree in combination with the green coverage degree of multiple images after the green coverage degree is obtained and calculating the leaf area index according to the comprehensive coverage degree and the plant height.

The invention discloses an image processing method, system and shooting terminal for optimizing edge aliasing. The method herein comprises the following steps: conducting smooth processing on a contour of which an image is to be processed to obtain a smooth closed contour curve, using the closed contour curve to conduct morphological reconstruction and Gaussian Blur processing on the image to be processed to obtain a fuzzy image of a binary image, then conducting a one-time gray scale statistics and threshold value processing on all the edges of the fuzzy image to obtain a first gray scale mapping table, and in accordance with the first gray scale mapping table, calculating so as to obtain a gray scale image, and further conducting a second gray scale statistics and threshold value processing on the aliasing edge of the gray scale image to obtain a second gray scale mapping table, and in accordance with the second gray scale mapping table, adjusting the aliasing edge to obtain an anti-aliasing image. According to the invention, the algorithm is simple, the operation speed is fast, and the improvement on the aliasing edge is explicit, and the application range is wide.

The invention discloses a single-image-based global depth estimation method, which comprises the following steps of: 1) performing Gaussian blurring on a single original image to be processed to obtain a blurred image; 2) converting the original image and the blurred image from a red, green and blue (RGB) three-channel space into a single-channel gray value space respectively; 3) acquiring a grayvariance value of each pixel in the original image and the blurred image according to gray values respectively; 4) calculating a variance ratio corresponding to each pixel in the original image according to the gray variance values; 5) performing normalization operation on a depth information value corresponding to each pixel in the original image, and performing conversion to obtain an initial global depth map of the original image; and 6) bilaterally filtering the initial global depth map obtained by the step 5) to obtain a final global depth map of the original image. By the single-image-based global depth estimation method, information in a global range can be effectively obtained, so that the accuracy of the obtained final global depth map can be ensured.

The invention belongs to the image processing field and includes a face key point-based automatic under-eye bag removing method and system. The method includes the following steps that: an input image is obtained; the masking-out image and under-eye bag area image of the under-eye bag area of the input image are extracted; the under-eye bag area image is transformed to an LAB color space, and a color channel L, a color channel A and a color channel B are separated out; a brightness adjustment curve is obtained through calculation based on the channel L; the channel L is mapped by the brightness adjustment curve, so that a new channel L is obtained; the new channel L, the channel A and the channel B are converted into an RGB color space image; Gaussian blur is performed on the masking-out image of the under-eye bag area, so that a new masking-out image can be obtained; the under-eye bag area image, the RGB color space image and the new masking-out image are composited, so that a new under-eye bag area image can be obtained; and the under-eye bag area image is replaced by the new under-eye bag area image, so that a final image can be obtained. With the face key point-based automatic under-eye bag removing method and system of the invention adopted, the removal of under-eye bags can be completed through one-key operation. The method is simple and stable, and the effect of an image processed by using the method is natural.