38results about How to "Reduce comparison" patented technology

The present invention relates to a device and method for iris recognition capable of improving the accuracy of iris recognition, which involve considering the reaction of pupils to the brightness of light and the resulting variations in the iris regions. The device and method are configured such that a plurality of iris images are captured and obtained using a camera and the brightness of light is adjusted from the point that the size of the iris is at a maximum up to the point that the size of the iris is at a minimum, the plurality of iris images are registered and stored, together with iris size information corresponding to each of the iris images, in a database interoperating with the iris recognition device, and an iris image for authentication, obtained for authentication or identification purposes using a camera, is compared with an iris image selected from the plurality of registered iris images, wherein the iris size of the selected iris image is most similar to that of said iris image for authentication.

Embodiments disclosed herein can significantly optimize a root cause analysis and substantially reduce the overall time needed to isolate the root cause or causes of service degradation in an IT environment. By building on the ability of an abnormality detection algorithm to correlate an alarm with one or more events, embodiments disclosed herein can apply data correlation to data points collected within a specified time window by data metrics involved in the generation of the alarm and the event(s). The level of correlation between the primary metric and the probable cause metrics may be adjusted using the ratio between theoretical data points and actual points. The final Root Cause Analysis score may be modified depending upon the adjusted correlation value and presented for user review through a user interface.

The invention belongs to the technical field of image processing, and discloses an image compression intra-frame coding prediction method and hardware implementation thereof. The image compression intra-frame coding prediction method comprises the following steps: acquiring an image to be compressed, and partitioning the image into L macroblocks; acquiring a brightness value of each pixel in an lth macroblock, and partitioning a brightness matrix of the lth macroblock into 16 4x4 brightness sub-matrixes; acquiring brightness values of 13 elements surrounding a pth brightness sub-matrix; acquiring 9 prediction modes, and calculating a sum of absolute difference of each valid prediction mode; selecting a prediction mode with a minimum sum of absolute difference from the valid prediction modes to serve as an optimal prediction mode of the brightness sub-matrix; adding 1 to a value of p to obtain optimal prediction modes of the 16 4x4 brightness sub-matrixes of the lth macroblock respectively; and adding 1 to a value of l to obtain optimal prediction modes of the 16 4x4 brightness sub-matrixes of each macroblock in the L macroblocks of the image to be compressed respectively.

The invention relates to a fast-moving star tracking matching identification method of a star sensor and relates to the field of satellite autonomous navigation technologies. By utilizing the fast-moving star tracking matching identification method, the problems of small number of tracking matching fixed stars obtained by the existing identification method of tracking star of the star sensor and poor stability of a tracking method are solved. The existing method has the problem of mismatching caused by the fixed star just entering the field of view. For ensuring matching effectiveness simultaneously, the number of the tracked matching fixed stars is further increased so as to improve the stability of a star tracking algorithm. In the matching process, in order to make a smaller amount of calculation, matching number state identification and borderline state identification arrays are subjected to operation with bitwise, and if a result is 1, matching is considered to be successful. By adopting the identification method with first sorting and second two-way recursive matching, pointless comparison is reduced, mismatching caused by the star newly entering the field of view is eliminated, most important of all, the number of the tracked matching fixed starts of every step is greatly improved, and the stability of the tracking algorithm is further improved.

The invention discloses a method for screening a nearest neighbor by using a potential neighbor relation graph in a recommendation system. The method comprises the following steps: (1) generating an object cluster set C with a redundant feature; (2) constructing the potential neighbor relation graph corresponding to the cluster set C; (3) quantifying weight of each edge in the potential neighbor relation graph, wherein the weight of the edge represents the possibility that two adjacent objects of the edge become the nearest neighbors; (4) clipping a potential neighbor relation to weed out redundant comparisons; (5) screening the nearest neighbor of a target according to the clipped potential neighbor relation graph. According to the screening method, on the premise of guaranteeing the recommendation precision, a recommendation based on a complete large-scale data set is mapped to a recommendation of a data set with a relatively small scale, so that the scale of the recommendation system is reduced, and high efficiency of the recommendation method is guaranteed.

A low-complexity column layered LDPC decoder implementation method, which adopts an efficient outer information compression and storage method on the basis of conventional LDPC layered decoding and compensates for the minimum and sub-minimum values ​​of the loss. , in the decoding process, each check node only needs to store the information tuples consisting of the minimum value and the second smallest value of the external information, which effectively reduces the storage resource requirements of the decoder for external information during the decoding process, and greatly reduces the The number of comparisons and replacements required for compressing storage computing is reduced, and the method can maintain excellent decoding performance while reducing storage and computing resources.

The invention provides a method for searching for webpage incompatible labels in a parallel mode based on MapReduce. The method is used for fast locating the incompatible labels. The method enables codes to be scanned to be divided into n blocks, a Map function is used on m computers for searching for the incompatible labels in the code blocks, and a used incompatible label base is a Trie structure organization. When the incompatible labels are found, the line numbers, the column numbers and the code block numbers of the labels are recorded. Finally, gathering processing is carried out on recorded results though a Reduce operation, and the positions of all the incompatible labels are returned to a user. The method can be used for the multi-platform deployment of a large-scale Web application programme, the incompatible labels can be fast located, an engineer is assisted in fast locating the incompatible labels needing to be processed in source codes, and deployed Web application is made to normally displayed on different browsers.

The invention discloses a column aggregation method of HBASE. Based on the HBase server query mechanism, the system architecture of the method is provided with an aggregation scanner module, and the description information of the aggregation table is expanded, and column aggregation attributes are added therein; in the query When merging, the multi-version data of the same cell is aggregated and calculated according to the column aggregation attribute to realize the aggregation on the cell; when the data is merged, the HBase server will have multiple versions according to the column aggregation attribute information in the aggregation table description. The data is aggregated and the result is retained, and the non-result data is cleared after the merge operation is completed. Compared with the prior art, the method disclosed in the present invention has better improvements in terms of real-time performance, performance and resource usage of aggregation table query.

The invention provides a data smoothing method and an apparatus thereof. The method comprises the following steps: A. distributing version numbers for an item of an upper layer module and an item of a lower layer module in a standby board by a main board, wherein items with same content in the upper layer module and the lower layer module have a same version number; B. when main / standby switching happens, switching the standby board to a new main board which executes a data smoothing process which comprises steps of C, D and E: C. to each item of the upper layer module in the new main board, according to a keyword of the item, searching an item comprising the keyword in a lower layer module of the new main board, if the item is searched, executing the step D, otherwise executing the step E; D. comparing whether the item and the searched item have a same version number or not, if not, using the item and the version number of the item to update the item searched in the lower layer module of the new main board and a corresponding version number, if so, ending processing of the item; E. downloading the item and the version number of the item to the lower layer module of the new main board.

The invention discloses a Chinese PINYIN quick word segmentation method based on a word search tree. The method is implemented by a computer or embedded mobile equipment and comprises the following working steps of: 1, building a Chinese character PINYIN search tree according to all the known Chinese character PINYIN lists; 2, combining the search tree with a hash table according to the built word search tree, and segmenting a string of given Chinese PINYINs; 3, working out a word segmentation result; and 4, destroying the search tree and releasing resources. Due to a public prefix of a character string, a construction space is saved, so that unnecessary character string comparison is greatly reduced; by the redundancy hash table with an index, the search efficiency is improved; and the time complexity of an algorithm is reduced to the minimum.

The invention discloses a method for recommending e-commerce commodities based on big data and multiple tags. The recommendation method includes the following steps: Step S1: Collect the search keyword of the commodity input by the user as the first keyword, and collect the purchased products in the user's history. The keyword of each product is the second keyword; step S2: judge whether there is a second keyword and the first keyword as matching keywords, if there is a second keyword and the first keyword as matching keywords, set the last The product corresponding to the second keyword is the product to be matched, and the product searched by the user inputting the first keyword is the search product, and the search product is marked according to the sales page content of the product to be matched and the search product; Step S3: from the marked search The products selected from the products are recommended to users first.