33results about How to "Keep the edges" patented technology

The invention belongs to the technical field of architectural ceramics, and specifically discloses a ceramic dry granular glaze, crystal diamond flashing ceramic brick and a preparation method thereof. The raw materials of the ceramic dry glaze include zirconium-containing transparent frit dry particles and zircon sand, and the zirconium source of the zirconium-containing transparent frit dry particles is zirconium oxychloride. The crystal diamond flashing ceramic tile includes a body layer and a surface glaze layer, and the surface glaze layer is fired from ceramic dry glaze. The present invention uses zirconium-containing transparent dry granule glaze as the basic glaze, and by adding zircon sand particles, because there is a small amount of evenly distributed zirconium in the zirconium-containing transparent dry granule glaze, the zirconium in the glaze melt is in a molten and saturated state, which can effectively prevent The high temperature melting of the extra zircon sand particles in the glaze melt. At the same time, the zirconium silicate crystals formed during the cooling process of the zirconium-containing transparent dry glaze use zircon sand particles as the crystallization carrier, so that the zircon sand particles grow further, and the crystal shape is more angular, making the prepared ceramics Dry-grain glazes can exhibit a diamond-like sparkle effect under light conditions.

The invention relates to an image denoising method based on a Shearlet contraction and an improved TV model, wherein a TV-denoising model is improved and a novel mixed denoising method by combining the Shearlet contraction is proposed. The method organically combines the sparse representation capability of Shearlet for a high dimension function with the protection capability of the TV-denoising model for an edge, wherein the method obtains a first-denoising image through a hard threshold function contraction, and then improves fidelity terms of a total variation model, and then makes a second denoising of the false Gibbs effect of the first-denoising image by combing the improved total variation model. On the prerequisite of protecting important information such as edges, etc., the method effectively inhibits the false Gibbs oscillation caused by the Shearlet contraction, and realizes a better visual effect and a lower computation complexity.

The present invention proposes an image denoising method based on superpixel clustering and sparse representation, which is used to solve the technical problems of low peak signal-to-noise ratio and loss of detail information of denoising images existing in existing image denoising methods, and the implementation steps 1. Input an image to be denoised; 2. Perform superpixel segmentation and superpixel clustering on the image to obtain multiple clusters of similar superpixels; 3. Perform image block extraction and dictionary training for each cluster of similar superpixels; 4 . Calculate the sparse coefficient of each image block under the corresponding dictionary; 5. Find similar image blocks of each image block, and calculate the weighted sum of sparse coefficients of similar image blocks; 6. Utilize the weighted sum of sparse coefficients of similar image blocks, Constrain the sparse decomposition process of each image block to obtain new sparse coefficients; 7. Determine whether the current iteration number is greater than the maximum iteration number Λ, if so, perform step 8, otherwise, increase the number of iterations by 1, and perform step 5; 8. Repeat Construct the image to be denoised to obtain the denoised image.

The invention discloses an oral-lip image automatic segmenting method based on Chinese medical inspection, comprising the specific steps as follows: (1) inputting a face image and carrying out mean shift filtering on the input face image to obtain an image I; (2) carrying out color space transformation on the image I to obtain a gray level image G and carrying out normalization processing to obtain a normalized image G; (3) obtaining a contour of oral-lip segmentation; (4) establishing a level set function according to the oral-lip image contour, iteratively calculating the level set functionand moving the contour to the oral-lip edge; and (5) setting the level set function to be 0 to obtain the oral-lip image contour, processing the oral-lip image contour by utilizing a five-point average method and eliminating burrs on the edge of the oral-lip image contour to obtain a final oral-lip contour. By means of the method, the oral-lip image can be automatically segmented from the face image, and the segmentation effect of a weak edge image can be effectively enhanced.

The invention relates to a fusion method of SAR (Synthetic Aperture Radar) images and visible light images on the basis of NSCT (Non Subsampled Contourlet Transform). The fusion method is characterized by comprising the following steps of: firstly, carrying out NSCT decomposition on the SAR images and the visible light images respectively; then adopting different fusion rules to carry out fusion treatment on NSCT low-frequency and high-frequency subband coefficients, wherein according to the decomposition coefficient characteristics of noise and signals in an NSCT domain, carrying out hard-threshold denoising on the NSCT high-frequency subband coefficient of the SAR images under the maximum decomposition scale, then respectively adopting different fusion rules to carry out fusion processing on the NSCT high-frequency subband coefficient under the maximum decomposition scale and other decomposition scales by adopting the coefficients with threshold processing as the basis; and finally,carrying out NSCT reverse transformation on the fused NSCT coefficients and obtaining fused images. The fusion method takes denoising as the basis of the fusion rule design, considers noise suppression while fusion treatment is carried out, is simple and easy to operate, can be used for obtaining a good fusion effect and is especially more applicable to the SAR images and the visible light imageswith serious spot and noise pollution.

The invention discloses a filtering noise reduction system and a filtering noise reduction method for video data. The method comprises the following steps: combining a mean filtering algorithm and a time-domain recursive filtering algorithm; generating a 5*5 neighborhood template through a neighborhood template generation module; dividing the sum of video image data of five columns of the 5*5 neighborhood template by a constant factor, so as to obtain a frame output; replacing data of an original frame by the mean of data of adjacent 25 frames in space so as to effectively filtrate noise caused by data stack in frames; allowing video image data subjected to mean filtering processing to be subjected to time-domain recursive filtering; carrying out subtraction operation between two adjacent frames; and multiplying the calculated result by a filtering coefficient, and then adding the second result with a current frame data so as to obtain the data of the current frame subjected to filtering. Therefore, the method is suitable denoising for the video images of low-velocity moving objects; and the frame duration of image signal is subjected to time weighted average, and the image signal is barely affected, so that not only are the edges of the images well reserved, but also the noise between frames caused by data stack can be eliminated.

The invention discloses a method and device for detecting changes between images, and relates to the technical field of image processing. The purpose is to detect changes between images easily affected by sensor noise and affect detection results when the monitoring area does not change under low illumination conditions. accuracy. The method includes: performing a morphological filtering operation on target images in the image set to obtain filtered images; using a compressed logarithmic ratio operator and a mean ratio operator to process the target images in the image set respectively to obtain the target image of each target image difference map, and perform equal-weight fusion operations on the difference maps of multiple target images to obtain a fusion difference map; compress the fusion difference image through a preset processing function to obtain a normalized fusion difference map, and normalize the fusion difference map The median filtering operation is performed to obtain the final difference map; the final difference map is clustered by a preset clustering algorithm, and the variation between images is determined according to the clustering result. The invention is suitable for detecting changes between images.