47results about How to "Preserve structural features" patented technology

The invention provides a method for extracting, matching and recognizing characteristics of finger veins, which comprises the following steps: acquiring a finger vein image by an infrared image acquiring device, preprocessing the image, extracting the characteristics, recognizing and analyzing, wherein the preprocessing step comprises the concrete steps of graying a color image, extracting a finger area, adopting direction filtering and enhancing, extracting a finger vein grain according to a finger outline mark and binarizing, de-noising by an area eliminating method, and standardizing the image size to a uniform image size; the characteristic extracting method comprises the steps of carrying out sub-block partitioning on the finger vein grain image, and extracting the characteristics ofeach sub-block image by a bidirectional two-dimensional principal component analysis method of bidirectional characteristic value weighting partitioning; and the recognizing and analyzing step recognizes the characteristics of each sub-block as a whole by a nearest neighbor sorter. The invention can reduce the calculated amount of a high dimensional image matrix in recognition of the finger veins, can obviously improve the recognition speed of the finger vein, and the recognition rate is stable and high.

The invention provides a finger vein feature extraction and matching recognition method. (1) Acquisition of finger vein images by an image acquisition device; (2) Preprocessing of the collected finger vein images, including: grayscale color images, finger area extraction, and combination filters to eliminate salt and pepper noise and Gaussian noise respectively Noise, use the local dynamic threshold algorithm to segment the image and binarize it, then use the area elimination method to denoise, extract the finger vein image according to the finger contour mark, and finally standardize the size of the image into a unified image; (3) A two-dimensional linear discriminant analysis algorithm weighted in each direction extracts finger vein features; (4) Matching and identification are carried out through the nearest neighbor classifier. The invention obviously improves the recognition speed of finger veins, and the recognition rate is stable and high.

The invention discloses an image impulse noise suppression method based on two-stage interpolation, which mainly solves the problem that the prior art can not retain image detail information in the de-noising process and also can not recover distinct images under the large noise density. The noise suppression method comprises the following steps of: carrying out noise detection and analysis on a noising image by utilizing a histogram and acquiring a low-resolution image through downsampling; carrying out first-stage interpolation by utilizing statistical properties of pixels without noise interference, completing amplitude information of pixels with impulse noise interference in the low-resolution image after the downsampling, and carrying out pre-filtering on the low-resolution image by adopting lifting wavelets on the self-adaption direction; and carrying out two-stage interpolation on the pre-filtered low-resolution image by adopting an improved segmental autoregressive model and utilizing the image structure characteristics to obtain a full-resolution image without the impulse noise. The invention has the objective evaluation and the subjective visual effect under different noise densities superior to the prior art, can be used for suppressing high-density impulse noise and improve the image quality.

The present invention belongs to the field of metrology, and more specifically relates to a spectral basic dimension correction and differential analysis method based on a moving least squares method.The method comprises: performing basic dimension correction on the spectrum obtained from known standard substances (series) by using the moving least squares method; using the corrected spectrum toperform differential analysis, specifically: using multiple spectra obtained from known standard substances (series) to perform cycle subtraction operations to obtain the maximum safety threshold of the system; performing cycle subtraction operations of multiple spectra between the known non-standard substances (series) and the known standard substances (series) to obtain the minimum warning threshold of the system; and performing cycle subtraction operations between the unknown sample spectrum and the known standard substances (series) spectrum, and obtaining a spectrum analysis result aftercomparing the difference values with the maximum safety threshold and the minimum safety threshold.

The present invention relates to a porcelain stoneware element (1) for the construction of driveways.

The invention discloses a method for predicting user characteristics based on user purchase behavior, the method comprising: collecting target user characteristic information, historical order information, and order commodity information, and using the order commodity information to obtain commodity-related triples from a public knowledge map Knowledge, constructing knowledge subgraphs, using the graph convolutional neural network to aggregate the local features of the entity's neighbors, and fully learning the representation vector of the entity; in the user feature prediction model, the target is learned according to different product features and different features of users with similar purchase behaviors The similarity between users and different commodities and users with similar purchase behaviors is fully based on the similarity between users and user entities to learn their feature vectors to meet the personalized needs of users. The present invention improves the accuracy of user feature prediction, so that user features can be predicted more accurately and a more complete user portrait can be constructed.