542results about How to "Improve time and efficiency" patented technology

The invention discloses an automatic collecting method of high-resolution satellite remote sensing traffic flow information. The automatic collecting method comprises the following steps of: A. pretreatment: registration of panchromatic images and a vector road network as well as panchromatic and multi-spectral images, road region division and double edge filtering strengthen; B. acquiring a vehicle sample characteristic value by visually judging a road region image obtained in the step one, and establishing a vehicle remote sensing image feature library; C. carrying out rough neural network vehicle extraction and fine facing objective vehicle extraction on the panchromatic images obtained in the step two; D. by utilizing a matching method relative to an image frequency domain, searching the vehicle position in fine extraction in the step three in multi-spectral images and carrying out matching; E. calculating a displacement amount of a same vehicle in the panchromatic and multi-spectral images according to the corresponding vehicle position obtained in the step three and the step four, thus estimating traffic flow parameter information; and F. verifying the traffic flow parameter information through precision evaluation. With the adoption of the method, static and dynamic traffic flow information in a large range series can be automatically and rapidly collected, the efficiency is high, and the method is simple and practicable.

The invention discloses establishment of a virtual case library of cancer pathological images and a multi-scale cancer detection system based on a convolutional neural network. The system is based ona method of a convolution neural network, a cancer mass region is detected on a pathological full-scan section, and the system includes four modules: 1) a pathological section image preprocessing module; 2) a virtual case database construction module; 3) a high-scale cancer mass detection module; and 4) a small-scale cancer mass classification module. The multi-scale cancer detection system provided by the invention can make full use of the multi-scale information of a pathological image, on different scales, according to the characteristics of the image, different strategies are designed to detect a suspected cancer area, and at the same time, under the condition of insufficient training data, the virtual case library method established by the invention can provide more training data setsfor an existing data-driven deep learning method. The multi-scale cancer detection system based on a convolutional neural network has the characteristics of multi-scale detection, driving of a relatively small amount of data and the like, and has the characteristic of reducing computing resources required for one-time recognition and improving time efficiency of an algorithm on the basis of ensuring the overall recall rate and accuracy.

The invention provides a method for detecting and identifying a station logo. The method comprises the following steps of: 1) setting a station logo example video frame, performing edge detection and filtering on the station logo example video frame, further positioning and segmenting a station logo area, and partitioning the station logo area to obtain a station logo template; 2) setting a video frame to be detected and identified, performing edge detection on the video frame to obtain an edge image of the video frame to be detected and identified; and 3) acquiring a station logo detection and identification result by a sliding window searching and block matching method according to the station logo template and the edge image of the video frame to be detected and identified. Compared with the conventional global matching-based method, the method of the invention can achieve higher station logo detection and identification accuracy rate and time efficiency.

The invention relates to a target retrieval method based on a group of randomized visual vocabularies and context semantic information. The target retrieval method includes the following steps of clustering local features of a training image library by an exact Euclidean locality sensitive hash function to obtain a group of dynamically scalable randomized visual vocabularies; selecting an inquired image, bordering an target area with a rectangular frame, extracting SIFT (scale invariant feature transform) features of the inquired image and an image database, and subjecting the SIFT features to S<2>LSH (exact Euclidean locality sensitive hashing) mapping to realize the matching between feature points and the visual vocabularies; utilizing the inquired target area and definition of peripheral vision units to calculate a retrieval score of each visual vocabulary in the inquired image and construct an target model with target context semantic information on the basis of a linguistic model; storing a feature vector of the image library to be an index document, and measuring similarity of a linguistic model of the target and a linguistic model of any image in the image library by introducing a K-L divergence to the index document and obtaining a retrieval result.

The invention discloses a random sampling consistency-based characteristic line detection method for a three-dimensional point cloud. The method comprises the following steps of: detecting a plurality of planes in the point cloud based on random sampling consistency; and applying the random sampling consistency to boundary points, which are taken as candidate points, of each plane parameterization area so as to obtain a final characteristic line. Therefore, the influences of noise, exterior points and data loss are reduced effectively, the detection robustness of the characteristic line is enhanced greatly and time efficiency is improved.

The invention relates to a low-temperature scouring and bleaching assistant and a preparation method and application thereof. The low-temperature scouring and bleaching assistant is a mixed assistant formed by 25-60 percent by weight of low-temperature oxidation bleaching activating agent and 40-75 percent by weight of matched scouring agent, wherein the low-temperature oxidation bleaching activating agent consists of sodium nonanoyloxy benzene sulfonate, an activating agent promoter, a bleaching promoter and ammonium polycarylate; and the matched scouring agent consists of an emulsifying agent, a penetrating agent, a chelated dispersing agent, a defoaming agent and water. Cellulose fiber unfinished textiles treated by the low-temperature scouring and bleaching assistant have better whiteness, capillary effect, strength and hand feeling than the performances of textiles treated by the traditional process. Compared with the traditional method, the invention can effectively save the consumption of steam and power, reduce the treatment pressure of sewage and reduce the treatment expense of the sewage.

The invention relates to a method for re-identifying persons on the basis of deep learning encoding models. The method includes steps of firstly, encoding initial SIFT features in bottom-up modes by the aid of unsupervised RBM (restricted Boltzmann machine) networks to obtain visual dictionaries; secondly, carrying out supervised fine adjustment on integral network parameters in top-down modes; thirdly, carrying out supervised fine adjustment on the initial visual dictionaries by the aid of error back propagation and acquiring new image expression modes, namely, image deep learning representation vectors, of video images; fourthly, training linear SVM (support vector machine) classifiers by the aid of the image deep learning representation vectors so as to classify and identify pedestrians. The method has the advantages that the problems of poor effects and low robustness due to poor surveillance video quality and viewing angle and illumination difference of the traditional technologies for extracting features and the problem of high computational complexity of the traditional classifiers can be effectively solved by the aid of the method; the person target detection accuracy and the feature expression performance can be effectively improved, and the pedestrians in surveillance video can be efficiently identified.

The invention discloses a quick binocular stereo matching method based on superpixel segmentation. The method comprises the following steps that: (1) adopting a SLIC (Simple Linear Iterative Clustering) superpixel segmentation method to carry out regional division on an original reference diagram and a target diagram; (2) on the basis of a local matching algorithm of adaptive weighting, calcuating an original parallax spatial diagram; (3) carrying out regional parallax plane fitting based on a confidence point; (4) applying a clustering algorithm to combine adjacent regional parallax planes; and (5) on the basis of a superpixel region, constructing an energy cost function to carry out stereo matching. A stereo matching unit is the superpixel region combined on the basis of the edge information of an image, so that the algorithm is guaranteed to be suitable for a large no-texture region, a depth boundary can be accurately positioned, a matching cost better than the matching cost of a traditional stereo matching method is obtained and is combined with a parallax postprocessing method to effectively obtain a high-accuracy parallax diagram, and the method exhibits good instantaneity.

The invention provides a method for optimizing laser radar detection atmospheric composition spectral line analysis, through building of an IPDA laser radar equation of hard target reflection, calculation of difference optical thickness of two pulses and CO2 air column mixing rate XCO2 with a weight, single spectral line broadening calculation, absorption cross section calculation, CO2 absorption optical thickness calculation, a weight function profile and a CO2 molecule number density profile, weight function-optimized spectral lines are obtained. The spectral lines obtained by the method have a higher lower-atmosphere weight ratio, and detection is closer to a true value.

The invention discloses a data stream anomaly detection system based on empirical characteristics and convolution neural network. The system includes an empirical feature extraction module, which is used to identify statistical features and header features as features based on artificial experience, which play a more important role in data packet anomaly recognition; a bit stream conversion picture module used to convert the data stream into the form of two-dimensional gray-scale picture, and then through the convolutional neural network perception, the global high-level perception features are extracted; a fusion splicing module used for fusing the above modules as the data stream characteristics and identifying abnormal data streams by using the full connection layer of the neural network; a distillation model module that replaces complex networks in actual deployment; a concept drift fine-tuning module updates the detection model of concept drift; an update experience database module adding new network attacks or hidden attack instructions to the artificial experience database. The invention accurately and efficiently detects abnormal behaviors such as network failure, user misoperation, network attack and the like.

The invention discloses a dynamic evaluation method of cell coverage in a communication network, which comprises the steps of: analyzing a plurality of measurement reports with different formats in a communication network to obtain measurement report data; computing a key coverage correlated vector between a target cell and any one measurement cell according to the measurement report data to obtain an orientation vector between the two cells; modifying the vector geographical position of the orientation vector by using a position characteristic point; ensuring a cell azimuth angle of the target cell; obtaining a target cell coverage distribution map according to the modified orientation vector and the evaluated cell azimuth angle; computing the maximum coverage distance and the coverage area of the target cell according to the target cell coverage distribution map; and performing over coverage judgment to the coverage of the target cell according to the maximum coverage distance and the coverage area of the target cell, and evaluating the coverage quality of the target cell. The method has the important significances that the service quality of a mobile communication network is improved, and the perception level of a mobile subscriber is improved.

The invention provides an integral image-based quick ACCA-CFAR SAR (Automatic Censored Cell Averaging-Constant False Alarm Rate Synthetic Aperture Radar) image target detection method, comprising the following steps of: (1) providing a G0 distribution-based self-adaptive global threshold CFAR pre-segmentation algorithm used for generating a target index matrix by combining the statistical property of data; (2) providing an integral image-based G0 distribution statistical parameter quick estimation method, wherein the statistical parameter can be calculated through simple operations such as addition and subtraction once 2-order and 4-prder integral images of an original image are obtained during the implementation of the method; and (3) giving out a basic implementation process of the ACCA-CFAR SAR image target detection method. Through the integral image-based G0 distribution statistical parameter quick estimation strategy provided by the invention, the time efficiency of the method can be greatly improved and the time complexity of the method is irrelevant to the size of a sliding window; and the requirement of the existing automatic target recognition (ATR) system on the treatment of large-scene data can be met to a great extent.

Provided is a high-precision full-automatic zero returning compensation type body axis system balance calibration system which comprises a resetting subsystem, a loading subsystem, a loading head initial positioning subsystem, a resetting measurement subsystem, a control subsystem and a software subsystem. In order to solve problems of high precision and full automation of a zero returning compensation type body axis system balance calibration system, the high-precision full-automatic zero returning compensation type body axis system balance calibration system with leading technology and advanced devices is developed. Experiences and lessons in ages with respect to the development and application of the balance calibration system are integrated into the development process of each subsystem so that the complete set of the system finally realizes the high-precision and full-automatic balance calibration.

The invention relates to a brillouin scattering spectrum real-time spectrum analyzing device and a data processing method thereof. The device comprises a low-pass filter, a low-noise amplifier, an analog-to-digital converter, an FPGA, a DSP and at least three memories, wherein the input end of the low-pass filter is connected with a brillouin scattering spectrum signal, the low-pass filter, the low-noise amplifier, the analog-to-digital converter and the FPGA are sequentially connected, all the memories and the FPGA are connected into a ping-pong structure, each memory circularly works according to the processes of storing time domain data->carrying out FFT processing-> outputting frequency domain data, which are sequentially and continuously carried out under the control of an FPFA, and the processed spectrum data is output in real time. The invention provides a high-speed data processing device for rapidly surveying and mapping BOTDR spectrograms in real time and a data processing method thereof; compared with the traditional method, the invention can greatly reduce the required time of surveying and mapping and provides the brillouin spectrum of an optical fiber to be measured in real time to capture the dynamic change information of strain or temperature.

The invention discloses one dosage type low temperature scouring and bleaching agent which comprises a surface active agent, a hydrogen peroxide activator, a bleaching synergist, a chelating agent, a hydrogen peroxide stabilizer, a defoamer and a filler in a mixing mode. The one dosage type low temperature scouring and bleaching agent can be used for scouring and bleaching cotton, cotton blend knitted fabric or cotton loose fiber in a low temperature bath of 60 DEG C- 85 DEG C and cotton or cotton mixed and blended yarns in a low temperature bath of 80 DEG C -95 DEG C and for cold pad-batch processing of cotton fabric. The one dosage type agent is a solid product which is convenient to pack and transport. One agent can replace more than three agents such as alternative scouring agent, hydrogen peroxide activator, chelating agent, brightening agent and oxygen bleaching stabilizer during processing, which is convenient to feed and process. The one dosage type low temperature scouring and bleaching agent is widely applicable to processes such as scouring and bleaching in bath and cold pad-batch processing.

The invention discloses an intelligent transportation moving object detection tracking method which is used for detecting and tracking a moving object in an intelligent transportation video. The intelligent transportation moving object detection tracking method comprises the following steps: considering objects with obvious edge characteristics such as guard bars in the middle part and on the two sides of a road in a video frame image; adopting a finite difference first-order partial derivative matrix to calculate the gradients of an input front video frame image and a back video frame image, and extracting a fine outline of an input image characteristic area; performing outline connection, area filling consolidation and morphological processing to obtain a preprocessed image; adopting an improved SIFT algorithm to extract key points from the preprocessed image, and only establishing a six-layered Gaussian pyramid for the preprocessed image, so as to ensure the situation that a few of key points can be detected when the image is changed in a small scale, meanwhile reduce redundant calculation requirement on establishing the pyramid, and improve the time efficiency of the algorithm; before frame difference, the image of the video sequence front frame and back frame is rectified by utilizing an affine transformation model obtained by the preprocessed image, so as to eliminate the influence of the video frame shift on performance detection.

The invention relates to an intelligent blind guiding device and a blind guiding crutch provided with the same. A central processing unit obtains a user location according to a positioning module, obtains destination information through a voice input module, then determines an optimal walking trip route through a path planning module, and plays navigation information through a voice prompt module; and road condition information is analyzed in real time during navigation with the combination of a data storage unit, an image acquisition module and road condition analysis module, and prompt information is played in real time through the voice prompt module to guide a user to travel. The intelligent blind guiding device is integrally arranged on the blind guiding crutch, hats and other articles for daily use, thereby being able to make accurate detection and reasonable evaluation of traffic signals, obstacles, potholes, pedestrians, vehicles and the like, provide timely and reasonable voice prompts for the user, provide convenience for travels of special population fields especially for the blind and the elderly, and achieve precise positioning, path navigation, traffic detection and voice prompt functions, so that the intelligent blind guiding device has a wide popularization significance.

The invention provides a passive video inter-frame forgery detection method based on the light stream consistency in the field of video content safety. The detection method mainly comprises the steps of extracting light stream features of two adjacent frames in a surveillance video shot by a static camera, adding light stream absolute values in horizontal and vertical directions to generate two light stream sequences of the whole video in horizontal and vertical directions, analyzing the light stream sequences, and performing video inter-frame forgery detection for features of different inter-frame forgery modes. According to the detection method, through the light stream features which are high-robustness video features, five types of video inter-frame forgeries, including video frame insertion, frame deletion, repeated duplication of single frames, non-adjacent duplication of continuous frames and adjacent duplication of continuous frames can be detected effectively, and a forceful detection weapon is provided for the integrity and the authenticity of court testimony videos.

The invention provides a duplicate video identification method and device and a terminal. The method comprises the following steps: extracting a key frame of a video to be identified; extracting a visual feature vector according to a key frame of the video to be recognized, wherein the visual feature vector comprises a plurality of visual eigenvalues; obtaining a key frame of a stored video corresponding to a visual eigenvalue from a key frame image retrieval database; calculating a video similarity between the video to be identified and the stored video according to the key frame of the videoto be identified and the key frame of the stored video; judging whether the video to be identified and the stored video are duplicate video according to the video similarity. The method avoids the tedious process of calculating the similarity of key frame sequences in pairs for a large number of videos, improves the time efficiency of video repetition detection, and improves the recall rate and accuracy of video repetition detection.

The invention relates to an important point-based time sequence fixed segmentation algorithm. According to the main technical characteristics, the algorithm comprises the steps of performing normalization processing on time sequence data; calculating segmentation fitting errors and performing sorting according to priorities of the segmentation fitting errors from high to low; extracting first k segments with maximum fitting errors and performing sorting from long to short according to segment lengths of a time sequence; extracting first two segments sorted according to the segment lengths, and performing fitting preprocessing; and comparing symbol and value relationships of maximum values and minimum values of the fitting errors in segmentation, determining important points to perform segmentation, progressively increasing the number of the important points, and performing loop iteration until a fixed segment number is reached. The algorithm is reasonable in design, greatly reduces the fitting errors, improves the segmentation efficiency, has very good fitting effect and adaptability, and can be widely used in the fields of time sequence prediction, clustering, anomaly detection and the like.

The invention discloses a perceptual hashing-based practical and safe image forensic system and a perceptual hashing-based practical and safe image forensic method. The system comprises a certificate authority (CA), forensics computing platform (FCP) and a forensic center (FC). The method comprises the following steps that: the CA provides a certificate for a user, the FCP and the FC and distributes and manages a control key of a hash function for the FCP and the FC; the FCP calculates a characteristic value and generates a hash value of an image at a registration or forensics stage; the FC manages an image hash value from the FCP and provides retrieval and forensics services; the user submits the image to the FCP, extracts an image retrieval characteristic by using the FCP, calculates the hash value of an image segment by using a key distributed by the CA and submits the hash value to an FC database; and when the user needs an authentication image, the image is also submitted to the FCP, the retrieval characteristic is extracted to retrieve the most coincident original image from the FC, the image is segmented and the hash value is calculated by using the same key and the difference between the image and the hash value is compared so as to determine a position on which content tampering occurs. The system achieves high balanced performance in the aspects of safety, robustness, sensibility, rapid retrieval and the like.

The present invention discloses a field detection control method of tomography foundation pile acoustic transmission method and device thereof, wherein the method comprises includes the following step: A. placing an acoustic emission sensor to the bottom of emission acoustic pipe and placing a gating acoustic receiving sensor to the bottom of receiving acoustic pipe; B. setting a control unit by the computer system CPU; C. detecting the nonholonomic detecting sector sections with the number of pile bases by the control unit; D. detecting M-2*J detecting sector sections at the middle segment ofthe pile body by the control unit K; and E. detecting J nonholonomic detecting sector sections by the control unit K. A depth position coder, an acoustic wave transmission sensor and the gating acoustic receiving sensor are connected with a sonic apparatus. The gating acoustic receiving sensor is composed of a gating component, a settable amplifying component, N detecting components, a control line L21 and a signal line L22. The method of the invention has the advantages of simple operation and simple structure and is suitable for the field detection for executing the acoustic transmission method for detecting the integrity of the foundation pile based on tomography.

The invention discloses an incremental learning and face recognition method based on locality preserving nonnegative matrix factorization (LPNMF), and relates to the technical field of pattern recognition. The incremental learning and face recognition method is essentially a face recognition method based on on-line learning. The method comprises the step (a) of face image preprocessing, the step (b) of initial sample training, the step (c) of incremental learning and the step (d) of face recognition. The incremental learning and face recognition method can be applied to a linear face recognition system and a non-linear face recognition system, keeps local structures of original space of face images, greatly improves recognition rate, and can be actually applied in an on-line mode.

The invention discloses a robust real-time on-line camera tracking method which comprises the following steps: estimating the posture of a camera and calculating and adding a key frame, wherein fast and general feature matching is adopted for the estimation of the posture of the camera, and the posture of the current frame can be estimated in a robust mode; and simultaneously tracking by means of the camera and establishing the scene mapping. By utilizing the robust real-time on-line camera tracking method, a more stable matching effect is generated under the condition of a larger scene or quick motion of the camera, and the limitation that the traditional camera tracking method is dependent on the partial matching is overcome. On the other hand, by utilizing the robust real-time on-line camera tracking method, the processing speed for the key frame is faster, so that more key frames can be contained in the scene, and the responding capability of a camera tracking algorithm for the problem that the camera is easy to fail in tracking when entering into an unknown scene is strengthened.

The invention discloses a fast movable object orbit clustering method based on sampling, belonging to the field of orbit data clustering. The method comprises the following steps: dividing each original orbit into an orbit dividing subset by utilizing a minimum length describing principle after a user sets an input parameter and sends a clustering analysis request; carrying out clustering analysis for the orbit dividing subset according to the similarity measurement among line sections to obtain an orbit clustering set; finally generating a representing orbit and a coverage area in a clustering way by each orbit, outputting and visualizing the representing orbit and the coverage area and returning a result to the user. The invention has the advantages that the method is used for the clustering analysis of data of a large-scale low-particlesize movable object orbit and can still keep the orbit clustering effect and also effectively discover the orbit clustering under the condition of very low particlesize of orbit data to be analyzed, thereby enhancing the system efficiency.

The invention discloses a K-means and deep learning-based image classification algorithm. The algorithm comprises the steps of 1) taking untagged images as input images, and randomly extracting image blocks to form an untagged image set of the untagged images same in size; 2) extracting a primary optimal clustering center by adopting a K-means algorithm; 3) constructing a feature mapping function, and extracting image features of the untagged image set; 4) performing pooling operation and normalization processing; 5) extracting a secondary optimal clustering center by adopting the K-means algorithm, extracting final image features by adopting convolution operation, and performing standardization processing on the final image features; and 6) classifying the final image features subjected to the standardization processing through a sorter. The algorithm has the advantages of simplicity, high efficiency, few training parameters and the like, and has a very good effect for classification of massive high-dimensional images; and the input images are preprocessed, so that the effects of improving the image classification effect and enhancing the classification precision are achieved.

The invention provides a road traffic monitoring radar and a realization method thereof. The radar radiates an emission signal to the external, the emission signal includes a signal segment A and a signal segment B, the frequency of the signal segment A linearly increases according to a set stepping value, the frequency of the signal segment B linearly increases according to the same stepping value, and the initial frequency of the signal segment A is different from that of the signal segment B; the radar receives a reflection signal, in response to the emission signal, of one or multiple target in the radar detection range; and the radar analyzes the frequency spectrum of an intermediate frequency signal which is obtained frequency mixing of the reflection signal of a random target and the emission signal, and calculates the distance and speed information of the target corresponding to the intermediate frequency signal according to the frequency and phase difference of the intermediate frequency signal. The distance, speed and orientation of the target can be obtained in one modulation period, and the time efficiency is improved; and when the targets are arranged densely, distance and speed information of each target can be accurately discovered.

The invention mainly relates to the technical field of quick satellite earth observation, in particular to a method for realizing hybrid and active scheduling on quick satellites. The method comprises the following steps: S1, generating an initial flexible scheduling plan: alternately performing heuristic search and constraint propagation until processing on all tasks is completed, wherein the resource allocation problem is solved by the heuristic search, and the activity start time range is reduced by the constraint propagation; S2, performing hybrid and active adjustment on the scheduling plan: performing scheduler underlayer processing by utilizing the minimum disturbance updating method of the constraint propagation when the task observation start time, the activity duration and the activity observation posture are adjusted; performing adjustment on a task observation time window and insertion of tasks, which are not arranged, by utilizing the heuristic search method for inserted tasks. The method for realizing the hybrid and active scheduling on the quick satellites, disclosed by the invention, is simple in operation, high in flexibility and easy to realize, and can be applied under the complicated background that the multiple quick satellites are scheduled.

The invention discloses a human age estimation method based on self-adaptation sign distribution. According to the method, through a self-adaptation training method combined with the application of an age sign distribution model algorithm, the obtained facial image feature vectors and initial age sign distribution are used as inputs, the KL divergence between the input initial age sign distribution and model prediction sign distribution is minimized by applying the age sign distribution model algorithm, and therefore the prediction age sign distribution can be obtained. Then, the prediction age sign distribution of the same age can be used for studying the age sign distribution of the corresponding age, circulation is carried out until training convergence is achieved or the set largest number of times is reached, and then the final prediction model can be obtained. Facial image features to be estimated are input into the final prediction model, and then the prediction age can be output. The age estimation accuracy of the system can reach the level similar to that of the human.

The invention discloses a distributed file system based IO (input output) request dispatching method. The distributed file system based IO request dispatching method includes receiving an IO request from a client; updating an overall situation process table according to the IO request; adding the IO request to the end of a request line of an overall situation object which the IO request belongs to; judging whether length of the continuous IO request of the overall situation object is smaller than REQ_MAX (L) or notand the L is smaller than maximum series of the multi-grade request line , if yes, deleting the overall situation object from the Lth line, and inserting the same to the end of the Lth+1 line; continuously monitoring the IO request from the client, repeating the above steps to generate a dynamic multi-grade progress request line; extracting all the IO requests of a first progress object in the multi-grade progress request line which is not null currently, and executing and updating status messages and parameters of the progress object. By the method, overall response time of a client request can be greatly reduced when system load is bigger, and performance of the system can be improved.