568results about How to "Heavy calculation" patented technology

A method for fast macroblock mode decision is disclosed. The method includes: (A) determining if a motion cost at the origin (0, 0) or a prediction motion vector (PMV) for a 4n*4n macroblock is smaller than a first threshold; (B) if the motion cost is smaller than the first threshold, determining a macroblock mode as 4n*4n and ending the method; (C) if the motion cost is not smaller than the first threshold, using an adaptive diversity search strategy to perform motion estimation on four 2n*2n blocks associated with the 4n*4n macroblock; (D) determining if all motion costs of the four 2n*2n blocks in step (C) are smaller than a second threshold; and (E) if step (D) determines that the motion costs of the four 2n*2n blocks are smaller than the second threshold, determining the macroblock mode as 2n*2n and ending the method.

In one embodiment of the present invention, luminance information of a video signal inputted thereto is weighted by a weighting section. An image characteristic value determination section, by one frame, takes an average of the weighted luminance information, thereby determining an image characteristic value of one frame. A backlight controlling section sets luminance of the backlight at a backlight luminance determination section in accordance with the image characteristic value determined by the image characteristic value determination section and performs luminance correction.

A method of computing a cryptographic key to be shared between a pair of correspondents communicating with one another through a cryptographic system is provided, where one of the correspondents receives a certificate of the other correspondents public key information to be combined with private key information of the one correspondent to generate the key. The method comprises the steps of computing the key by combining the public key information and the private key information and including in the computation a component corresponding to verification of the certificate, such that failure of the certificate to verify results in a key at the one corespondent that is different to the key computed at the other correspondent.

A wireless access point is operable to tag packets received from mobile clients. For example, a VLAN tag or Layer-3 tunnel tag for a packet is determined based at least in-part on a mapping between a mobile client and tag maintained by the access point. In one embodiment the access point uses Association IDs (“AIDs”) to uniquely identify mobile clients in the BSS. Hence, the mapping is between A / Ds and VLAN tags / Layer-3 tunnel tags. The mapping may be generated by snooping authentication related communications or from information maintained by a switch, such as a MAC address to tag mapping.

A system and method for media adaptation are provided. A method for providing media to an electronic device includes receiving a request for a media with a set of characteristics, and determining if the media with the set of characteristics resides in a media storage. The method also includes if the media with the set of characteristics resides in the media storage, retrieving the media with the set of characteristics from the media storage. The method further includes if the media with the set of characteristics does not reside in the media storage, transcoding an alternate media with an alternate set of characteristics to produce a transcoded media with the set of characteristics. The method additionally includes providing the media with the set of characteristics from the media storage or the transcoded media with the set of characteristics to the electronic device.

A radio-wave arrival-direction apparatus calculates a correlation matrix of received signals by correlation calculation between antenna elements, and calculates a noise spatial eigenmatrix, of which each row or column is an eigenvector belonging to a noise eigen-space, by eigenvalue factorization of the correlation matrix. The apparatus also factorizes a matrix including a product of the noise spatial eigenmatrix and a conjugated and transposed matrix of it to an upper or lower triangular matrix, using cholesky factorization. The apparatus calculates an angle evaluation value in a predetermined angle range of an arrival-angle evaluation function using the derived upper or lower triangular matrix, and determines an arrival angle based on the calculation result. A calculation amount in a variable angle range can be thus reduced without causing accuracy degradation of arrival direction, in an algorism requiring all angle sweep for arrival angle estimation of MUSIC method or the like.

The invention provides a target tracking method and system and a staff behavior analyzing method and system. The target tracking method comprises the steps that a camera is used to acquire a video image; a front head classifier, a back head classifier, a left head classifier and a right head classifier are respectively used to detect candidate head targets in the current image; deduplication and screening are carried out on the candidate head targets detected by the classifiers; head targets which does not meet a tracking condition are deleted; and the remaining head targets are respectively tracked. The technical problems of false positives, false negatives, difficulty tracking and high camera erection requirement, which are caused by easily being blocked, difficulty target classification and the like, of a staff detection and tracking method based on face detection, head and shoulder detection or whole body detection are solved. According to the invention, accurate staff track tracking can be carried out based on the head targets; the problem of target restating can be reduced; and staff behavior analyzing can be carried out based on tracking results.

The present invention provides a Hybrid GA (HGA) in which local optimization operation is performed to only a few individual with lower fitness selected from each population of widely distributed generations in genetic algorithm (GA). Since this HGA shows very powerful global searching ability even in a vast multi-dimensional parameter space with strong multi-peak feature, the present invention may efficiently determine a number of structural parameters from a little information included in X-ray diffraction circles. As a result, the present invention may determine a complicated material structure from powder specimen even though it is hard to make a single crystal specimen. Thus, the development of new materials would be highly accelerated in the field of medicines or materials, etc.

The invention relates to the digital image processing field and particularly relates to a white balance and dark primary color adaptive histogram underwater image enhancement method. The method comprises steps that step 1, color correction for an image is carried out through utilizing a dynamic threshold white balance algorithm; step 2, a dark channel map of the image after color correction is solved through utilizing a dark channel model; step 3, a weight factor of block images is calculated; step 4, underwater image processing is carried out through utilizing a CLAHE method; step 5,the weight factor and gray mapping relationship tables are utilized to acquire a final gray mapping relationship table of a block through fusion calculation; and step 6, a bilinear interpolation algorithm is utilized to calculate a gray mapping value corresponding to each pixel in block images one by one, enhanced block images are acquired, interpolation calculation for pixels connected between the adjacent block images is carried out, and images of underwater images after enhancement are acquired. The method has advantages of simple calculation and strong timeliness.

Apparatuses, computer program product, and method for digital image processing. A digital image processing apparatus includes an input interface to obtain a first digital image and a second digital image, and a processing unit coupled with the input interface. The processing unit defines at least one block in the first digital image, defines for each block a search area in the second digital image, the search area being larger than the block, maps the block and its search area to an equal size, calculates pixelwise errors between each block and its search area that are mapped to an equal size, collects the errors into a motion register, and defines a motion between the first digital image and the second digital image by utilizing the motion register.

The invention discloses an iterative least square method-based MIMO (multiple input multiple output) radar DOA (direction-of-arrival) estimation method, which is characterized in that receiving and transmitting array response matrixes on which dimension-reduced processing is performed are solved by using an iterative least square method. The iterative least square method-based MIMO radar DOA estimation method comprises the following steps: firstly, performing the dimension-reduced processing on echo data matrixes of multiple radar transmitted pulses and the receiving and transmitting array response matrixes; then, establishing cost functions under the least square condition, and solving the cost functions by utilizing a gradient descent-based iterative method; finally, estimating the direction of a target by utilizing known receiving and transmitting array manifolds. Compared with a traditional monostatic MIMO radar array DOA estimation method, the iterative least square method-based MIMO radar DOA estimation method disclosed by the invention directly obtains the DOA estimation of the target, and does not need to perform spectrum peak search. Noise is effectively suppressed by adopting the dimension-reduced processing, and the estimation accuracy under low signal to noise ratio is improved; the estimation, the inversion and the eigenvalue decomposition operation of high-dimensional data covariance matrixes are avoided; the defects that the calculated amount is high and the needed sample number is large when the traditional array DOA estimation method is applied to a monostatic MIMO radar are overcome.

The invention discloses a clustering and reclassifying face recognition method, which comprises the steps of acquiring a training sample; carrying out equalization processing on the training sample; carrying out Gabor texture feature extraction on face images, and acquiring a feature vector corresponding to each face image after feature extraction; carrying out dimension reduction on acquired Gabor texture features of each face image to acquire feature vectors after dimension reduction; carrying out a clustering operation until distance convergence so as to complete clustering; classifying all of the clustered feature vectors to acquire a plurality of subclasses, calculating to determine each vector mean value, and calculating to acquire a within-class distance and an among-class distance; carrying out feature extraction and preprocessing on face images of a target to be recognized, acquiring a feature vector after projection transformation, and calculating the distance between the acquired feature vector and the feature vectors in each subclass sequentially so as to acquire the similarity; and determining identity information of the target to be recognized. The method disclosed by the invention can shorten the among-class distance so as to reduce an error in the acquisition process, and the accuracy of face recognition is improved.

The invention relates to the field of electronic information readers and discloses an electronic information reader and a display method thereof. In the invention, a sensor is used for monitoring the change of the effective display width of a flexible screen and carrying out renewal type-setting according to the current effective display width after a change occurs so as to ensure that a display mode is in optimal match with the effective part of the flexible screen without the occurrence of the situation of partially invisible content. When the change of the effective display width of the flexible screen is large enough or after the change is stopped, renewal type-setting and refreshment are carried out so as to reduce the frequency of type-setting and refreshment and the consumption of precious electric energy.

A system for performing a virtual colonoscopy includes a system for generating digital images, a storage device for storing the digital images, a digital bowel subtraction processor coupled to receive images of a colon from the storage device and for removing the contents of the colon from the image and an automated polyp detection processor coupled to receive images of a colon from the storage device and for detecting polyps in the colon image.

Te invention discloses a large-scale time-lag electric system characteristic value calculation method based on EIGD. The method includes the steps of establishing a time-lag electric system model, converting the formula of the time-lag electric system model into the abstract cauchy problem, converting the characteristic value of the time-lag electric system model into the characteristic value of an infinitesimal generator of the calculated and converted formula, conducting discretization on the infinitesimal generator to obtain an approximate matrix of the infinitesimal generator, obtaining the approximate matrix where displacement processing is conducted by conducting displacement processing on the approximate matrix, obtaining an inverse matrix after inverse conversion is conducted on the approximate matrix where displacement processing is conducted, converting the partial characteristic value required to be calculated into the partial characteristic value with the maximum module value, calculating the partial characteristic value, with the maximum module value, of the inverse matrix through the Arnoldi algorithm, and obtaining the characteristic value lambda of the approximate matrix. Verification is conducted through the Newton iteration method, and the accurate characteristic value and the accurate characteristic vector of a time-lag electric system are obtained through calculation.

Summarization of video content that includes the identification of a plurality of segments from a Video including baseball. In some embodiments the system does not use a model sequence of frames.

In a face recognition apparatus adapted to recognize an input face image on the basis of one or more registered face images, similarity between an input face image and a registered face image is determined based on a Gabor jet calculated by performing Gabor filtering using Gabor filters defined by a Gaussian function representing a window and sine and cosine functions representing a frequency response for each of predetermined frequency values and for each of predetermined angles of the response function. Values of the filter window are calculated in advance based on the Gaussian function and stored in a filter window data ROM. Values of the response function are calculated in advance for respective angles based on the sine and cosine functions and stored in sine data ROMs and cosine data ROMs. In the Gabor filtering process, coefficients of the Gabor filters are determined from values read from these ROMs.

Provided is a technology capable of efficiently saving data generated at an intermediate stage of an analysis processing and reusing intermediate data. Data generated at the intermediate stage of the analysis is saved, quantified feedback information for the saved data is received as an evaluation value, and the intermediate data that has not been given an evaluation value is preferentially deleted while the analysis processing for similar data is performed with regard to the intermediate data that has received a particularly high evaluation value, thereby performing automatic management of the intermediate data by a background processing so that the analysis of data to be subjected to a comparison and a derivatively-assumed analysis can be performed at high speed.

The related service gateway comprises: a message interceptor to transmit the intercepted user request and / or service response message to the next module and sends message from the next module to a message transmission module; a safe processing module to safe process message from last module and return result; and a message transmission module to transmit the processed message to a grid server. It also relates the corresponding processing method. This invention P2P transmits message safely, and realizes seamless visiting for service resource to mobile terminal.

The invention relates to a mobile phone information based section speed calculation method. The method is used for calculating the average speed information of a vehicle at each section through the analysis and the treatment of mobile phone information and the combination of traffic road position and direction. The key points of calculating the section average speed lies in massive data processing and precise positioning of a mobile phone user in a movement process. The problems are solved through two-stage three-time map matching in the invention. The map matching at the first stage comprises the following step of: matching Cell with sections, and the map matching at the second stage comprises the following steps of: establishing a subordination relationship of mobile phone data points and the sections through matching user data points with the sections two times, finally calculating the traffic speed of each user on the sections by using the Cell position in which continuous data points are located and the time, removing part of abnormal speeds in the traffic speeds, summing the traffic speeds and averaging the sum to obtain the average speed of each road. In the invention, the large amount of calculation caused by directly matching each data point with the sections is avoided, and the calculation efficiency is improved. The accuracy of the mobile phone data and the sections is improved through a proximity principle and a direction consistency principle.

The invention discloses a method for generating a random access pilot in low complexity in a long term evolution system, which is characterized by comprising the following steps of: directly generating a frequency domain sequence with the length of NZC; through frequency offset treatment, performing zero-padding on the frequency domain sequence to a nonequispaced fast Fourier transform NFFT point; through cyclic shift, performing inverse fast Fourier transform, performing K-time oversampling, adding prefix and the suffix and delivering to a filter; and through the frequency offset shift, finally adding a cyclic prefix to generate the random access pilot. Compared with the conventional methods, the method for generating the random access pilot in low complexity in the long term evolution system greatly reduces the complexity; meanwhile, the realized result of the method is consistent with that of the conventional methods, and is more suitable for an actual system.

The present disclosure discloses a GPU-based TFT-LCD Mura defect detection method, comprising: (1) establishing a studentized residual based double-second-order regression diagnosis model based original image data to obtain double-second-order regression background data; (2) obtaining influence quantities of respective data points on fitted values according to the original image data and the double-second-order regression background image data; (3) excluding outliers and influential points in the original image data according to the influence quantities to obtain a new pixel point set; (4) establishing a double-N-order polynomial surface fitting model according to the new pixel point set to obtain double-N-order background image data; (5) obtaining a residual image R according to the double-N-order background image data and the original image data, and performing threshold segmentation on the residual image to obtain a threshold segmentation image; and (6) performing morphological processing on the threshold segmentation image to obtain an eroded and dilated image, thereby achieving effective segmentation of Mura defects with uneven brightness distribution.

The invention relates to in an analysis method of the Facial Expression image classification and semantic evaluation on the basis of semantic evaluation and the quantization. This invention makes use of the sample collection of training Facial Expression images, extracting expression characteristics to form a marker map of LG Vector, projecting it to the Principal Component PCA subspace, making the use of these reduced-order LG Vectors to learn mixed multi-dimensional t-distribution, as the semantic judgment of six basic emotions of the images, in according to that which probability taken up by expression is maximal, that image is judged thereby to that expression. Which resolve the recognition difficulties of the uncertainties, automatic facial expression existed in the current technology, and the singularity difficult to overcome such defect. This invention is an extremely flexible and powerful modeling tools based on the statistics, provides a more robust ways, avoiding the extreme estimates of posterior probability of the subordinate component of sample observed, the training samples do not require labeling, do not require any post-processing, less sensitive to outliers, avoid this artificial Judgment.

The invention provides a system for realizing 360-degree panoramic video. The system is realized by being arranged on a camera module, and comprises an interested region selection module, an image stitching processing module and a video conference system module, wherein the interested region selection module is used for defining an arbitrary region within a horizontal plane 360-degree range coated with four lenses as an interested region, and performing image capture on the defined interested region; the mage stitching processing module is used for judging whether stitching operation is required or not according to images captured by the interested region selection module; the video conference system module is used for receiving a panoramic image transmitted by the image stitching processing module, and performing efficient coding processing on the image, so as to realize a 360-degree panoramic video conference effect. The system can randomly select interested regions to perform coding and transmission, and realize instantaneous switching of interested regions in the horizontal plane 360-degree panoramic image without camera mechanical operation.

A wireless access point is operable to tag packets received from mobile clients. For example, a VLAN tag or Layer-3 tunnel tag for a packet is determined based at least in-part on a mapping between a mobile client and tag maintained by the access point. In one embodiment the access point uses Association IDs (“AIDs”) to uniquely identify mobile clients in the BSS. Hence, the mapping is between AIDs and VLAN tags / Layer-3 tunnel tags. The mapping may be generated by snooping authentication related communications or from information maintained by a switch, such as a MAC address to tag mapping.

The invention belongs to the technical field of neural networks, and relates to a neural network model compression and acceleration method based on entropy attention. According to the invention, a large parameter is constructed; a teacher network model with a large calculation amount and excellent performance is used for carrying out strong supervised learning on a student network with a small parameter amount, a small calculation amount and poor performance, a small model with a small parameter amount, a small calculation amount and excellent performance is finally obtained through the learning process, and the small model can meet the requirements for real-time performance and precision of a real scene.

The present invention provides a display panel and a display method thereof, and a display device. The display panel comprises a plurality of circulation units, each circulation unit is composed of two rows of sub-pixels, the sub-pixels in each row include one red sub-pixel, one green sub-pixel and one blue sub-pixel, and the sub-pixels in a first row are provided to shift by a half size of the sub-pixel in a row direction with respect to the sub-pixels in a second row, respectively, and an arrangement mode of the sub-pixels in the first row is different from that of the sub-pixels in the second row. The display method comprises: determining initial components of the sub-pixels according to a picture to be displayed; and determining a display component of a sub-pixel according to the initial components of the sub-pixel and common sub-pixels of the sub-pixel.

The invention provides a display method and a display device. The display method in the invention is applicable to a display panel comprising multiple rows of sub-pixels, each row of sub-pixels are formed of sub-pixels of various colors which are alternately and circularly arranged, the sub-pixels in each row are arranged in the same order, and in column direction, sub-pixels of the same color are not adjacent, wherein the display method comprises steps of: S1, generating a primary image based on image information, the primary image is formed of virtual pixels arranged in a matrix, each virtual pixel is formed of sub-pixels of different colors and size of the virtual pixel is the same as that of the sub-pixel of the display panel; S2, calculating a display component of each sub-pixel by using primary components of sampling virtual sub-pixels of the sub-pixel.

The invention provides a display method and a display device. The display method comprises: generating a primary image based on image information, wherein the primary image is formed of virtual pixels arranged in a matrix, each virtual pixel is formed by multiple virtual sub-pixels of different colors arranged in one row, in the column direction, the size of the virtual pixel and that of the sub-pixel are the same; calculating a comparison component of each sub-pixel by using primary components of comparison virtual sub-pixels of the sub-pixel; acquiring differences between comparison components of each sub-pixel and other sub-pixels adjacent thereto, if each of the differences between comparison components of the sub-pixel and the other sub-pixels exceeds a predetermined threshold value, determining the sub-pixel as a highlighted sub-pixel; calculating a display component of each sub-pixel by using primary components of sampling virtual sub-pixels of the sub-pixel.

The invention relates to a binocular vision-based unmanned aerial vehicle aerial autonomous refueling fast docking navigation method and belongs to the machine vision and image processing field. The method includes the following steps that: a binocular camera system with an optical filter shoots a refueling taper sleeve with an optical marker lamp, so that a left image and a right image are obtained; gray processing, binarization and median-value filtering are performed on the left image and the right image sequentially, all communicated regions in the left image and the right image are searched and marked through adopting a regional growth method, and the centroids of each of the communicated regions are calculated and are adopted as feature points; improved haar wavelet transform is adopted to describe all the feature points, and the description vectors of all the feature points are obtained, the feature points of the left image and the right image are matched through adopting the minimum Euclidean distance method, so that the feature point pars of the left image and the right image are obtained; and the three-dimensional coordinates of the feature points are calculated by using a binocular vision principle, and the least squares method is adopted to perform spatial circle fitting on the three-dimensional coordinates, and the end surface circle curve of the refueling taper sleeve is obtained, and the center, normal vector and radius of the circle are obtained through calculation.