156 results about "Region analysis" patented technology

It is an object of the present invention to provide a method, an apparatus and a program having high optimization precision and capable of obtaining an answer required by a designer in a short time by combining optimization between individual transistors and optimization as the entire circuit, or by appropriately combining judgment of an operation region, an analysis of the operation region and a SWEEP sensitivity analysis when the optimization is carried out. An optimizing designing apparatus of an integrated circuit for designing a circuit, comprises operation region judging means for adjusting an operation region (linear region, saturation region) of the circuit, operation region analysis means for displaying liner characteristics (Ids-Vgs characteristics) of the circuit and saturation characteristics (Ids-Vds characteristics) of the circuit, and SWEEP sensitivity analysis means for displaying variation in output characteristics of the circuit.

Systems and methods for replacing original media bookmarks of at least a portion of a digital media file with replacement bookmarks is described. A media fingerprint engine detects the location of the original fingerprints associated with the portion of the digital media file and a region analysis algorithm characterizes regions of media file spanning the location of the original bookmarks by data class types. The replacement bookmarks are associated with the data class types and are overwritten or otherwise are substituted for the original bookmarks. The replacement bookmarks then are subjected to a fingerprint matching algorithm that incorporates media timeline and media related metadata.

Methods for feature selection based on hierarchical local-region analysis of feature relationships in a data set are provided.

The invention provides chip appearance detection method and system, which is applicable to the technical field of chip appearance detection, wherein the method comprises the following steps of: obtaining a picture of a chip to be tested; identifying whether a welding disc exists on the picture of the chip to be tested or not; storing the position information of the welding disc when the existenceof the welding disc is identified, and carrying out primary gray processing on the obtained picture of the chip to be tested according to the stored position information of the welding disc; carryingout communicated region analysis on the picture of the chip to be tested after the first gray processing; and identifying whether defects exist on the welding disc or not. The automatic identification of appearance defect of the welding disc on the chip can be realized, and the problems of low detection speed and low detection efficiency caused by welding disc defect of the manual eye detection mode can be avoided. In addition, after the step of obtaining the picture of the chip to be tested and before identifying whether the welding disc exists on the picture of the chip to be tested or not according to the shape characteristic of the welding disc, whether the dimension of the chip to be tested is qualified or not can be judged through extracting the dimension data of the chip to be tested and comparing the obtained data to the stored chip dimension data.

An adaptive automobile meter detection method based on character segmentation cascade classifier which processes threshold segmentation, morphology and connected components analysis on the original image; extracts the pointer based on the contour analysis method, establishes a pointer information list; constructs a character segmentation cascade classifier by combining a HOG / SVM character segmentation classifier, a character filter and a CNN digit classifier. The character segmentation cascade classifiers is used to identify the digit character area of the automobile meter. Region analysis is performed to extract tick marks based on the center of the digit character area. The angular position corresponding to the tick mark is determined. The response value corresponding to the pointer is determined by establishing the Newton interpolation linear description relationship between the pointer angle and the response value and then the meter is classified as pass or fail. The invention is suitable for the field of visual inspection of automobile meter pointers.

The invention discloses an image target detection method based on weak supervised learning, and belongs to the technical field of machine vision. The method comprises the following steps: firstly, collecting an image data set, and training a constructed deep convolutional neural network model based on a multi-scale feature map by adopting a multi-example learning method; inputting an actual image,and extracting a category thermodynamic diagram of the actual image through the deep convolutional neural network model; and finally, outputting a bounding box of a target in the category thermodynamic diagram by adopting a binarization image connected region analysis method to obtain a target detection result. According to the method, an image target detection task is realized based on a weak supervised learning method. The target detection task can be completed only by using the image-level classification annotation information in annotation in convolutional neural network model training, which is different from the target bounding box annotation information required in the prior art, so that the work of manually annotating the target in the image is greatly reduced, and the completionof the image target detection task has more economic benefits.

A method for feature selection based on hierarchical local-region analysis of feature characteristics in a data set of mixed data type is provided. A data space associated with a mixed-type data set is partitioned into a hierarchy of plural local regions. A relationship metric (for example, a similarity correlation metric) is used to evaluate for each local region a relationship measure between input features and a target. One or more relevant features is identified, by using the relationship measure for each local region.

The invention discloses an information locating method for a document image based on a standard template. The method comprises the following steps: first, binaryzation processing and connected region analysis are performed on an image to be processed to obtain a series of candidate connected domains, and a group of connected domains with the characteristic of easy recognition is searched among the candidate connected domains; then, tilt correction and perspective correction are performed on a gray image of the image to be processed according to the acquired connected domains with the characteristic of easy recognition; and finally, the position of information to be recognized in the corrected image is located according to the positions of the connected domains with the characteristic of easy recognition. According to the invention, a locating method is provided by directly making use of image recognition, the acquisition process is simple and requires neither special equipment nor design of an exclusive bar code, and the method has a wider range of applications.

The invention discloses a method for automatically extracting a gesture candidate region in a video sequence. The method comprises the following steps of: enabling a gesture video image acquisition system to acquire video images; constructing reference background images; computing to generate motion description images; computing to obtain motion segmentation threshold values and converting the motion description images into binary motion images; conducting skin color segmentation to obtain binary skin color images; conducting logic and operation to the binary motion images and the binary skin color images point by point to obtain binary fused images; and conducting connected region analysis to the binary fused images and selecting the gesture candidate region. The method combines motion information with skin color information to find the gesture candidate region, the two kinds of information complement each other, the detection accuracy is improved, the method is practical and effective and a very good foundation is provided for the segmentation, the positioning and the recognition of gestures.

A method of automatically correcting dust artifact regions within images acquired by a digital image acquisition system including a digital camera with a lens assembly includes determining probabilities that certain pixels within multiple digitally-acquired images correspond to one or more dust artifact regions. A statistical dust map is formed including mapped dust artifact regions based on the dust artifact determining. The mapped dust artifact regions are separated into aura regions partially obscured by dust and shadow regions substantially obscured by dust inside the aura regions. Digital data are corrected corresponding to pixels within one or more digitally-acquired images including dust artifact regions separated into aura regions and shadow regions correlated with mapped aura and shadow regions of the statistical dust map.

The invention relates to a man-machine interaction method based on analysis of interest regions by bionic agent and vision tracking, comprising the steps: (1) a designer carries out user analysis and designs interest regions which can be possible to cause the attention of a user according to the user analysis result; (2) an event interaction manager receives and analyzes data generated by an eye tracker in real time, and calculates the focal positions of the eyeballs of the user on a screen; (3) the event interaction manager analyzes the interest regions causing the attention of the user according to the obtained focal positions of the eyeballs of the user on the screen; and (4) the event interaction manager takes the analyzed result of the interest regions causing attention of the user as a non-contact instruction to control the expressions, actions and voices of the bionic agent on the man-machine interaction method so as to carry out intelligent feedback on the user and further realize natural and harmonious man-machine interaction. A man-machine interaction system which is established according to the invention based on analysis of interest regions by bionic agent and vision tracking, comprises (1) the eye tracker; (2) a man-machine interaction interface; (3) the event interaction manager; and (4) the bionic agent.

The invention provides a low-voltage platform region intelligent management method based on a forward-backward substitution algorithm. The low-voltage platform region intelligent management method comprises the following steps: 1, determining a platform region with a low-voltage phenomenon; 2, collecting original data; 3, drawing a picture and setting electric element parameters; 4, carrying out load flow calculation; 5, analyzing low-voltage factors; and 6, carrying out low-voltage management and economic evaluation. According to the low-voltage platform region intelligent management method, a position with the low-voltage phenomenon in the platform region is found by algorithm calculation according to front end voltage and tail end voltage of the platform region; one set of intelligent management measures with best economical efficiency can be generated by combining an actual low-voltage phenomenon solving measure; and effective helps for platform region analysis and project approval are provided for field personnel.

The invention provides a Harris angular point and stroke width based text region detection method. The method mainly comprises such three parts of edge enhanced Harris angular point detection, stroke width based candidate region scanning and text color based region expansion. According to the invention, a text region is acquired through detection for angular points, connected region analysis is carried out, and a response region which does not conform to text line characteristics is removed, thereby being capable of significantly improving the accuracy of angular point detection for the text region in an image. Compared with a text detection method singly based on edge characteristics, the method provided by the invention can improve the recall rate of text detection in the image. Compared with a text region detection method singly based on connected regions, the method provided by the invention can acquire higher detection efficiency. The text region detection method has advantages when being compared with the text detection method singly based on the edge characteristics and the text detection method singly based on the connected regions.

A method of automatically correcting dust artifact regions within images acquired by a digital image acquisition system including a digital camera with a lens assembly includes determining probabilities that certain pixels within multiple digitally-acquired images correspond to one or more dust artifact regions. A statistical dust map is formed including mapped dust artifact regions based on the dust artifact determining. The mapped dust artifact regions are separated into aura regions partially obscured by dust and shadow regions substantially obscured by dust inside the aura regions. Digital data are corrected corresponding to pixels within one or more digitally-acquired images including dust artifact regions separated into aura regions and shadow regions correlated with mapped aura and shadow regions of the statistical dust map.

The invention discloses a multi-target detection and tracking method for a speed dome camera control strategy. The method mainly comprises the following steps: preprocessing acquired bolt video image data, removing surplus frames of image, carrying out foreground point detection on the [2, N] frame of image, updating a background model, carrying out communicated region analysis, computing target characteristics, computing and comparing similarity of targets in adjacent frames, carrying out multi-characteristic similarity fusion, carrying out optimal matching, finding out targets, updating states and updating target characteristics. The method has the advantages of improving the accuracy of target tracking, dealing with separation and integration events very well and finding out target IDs related to the events; in addition, the similarity accuracy can be effectively improved so that the precision in the matching process is improved, the matching accuracy is enhanced greatly and the mismatching rate is lowered greatly; the IDs of targets after integration and separation can be found again accurately; the algorithm real-time property is high, the operation efficiency is very high, the required memory is less, and therefore, the method can be implemented in a front-end embedded environment.

The invention discloses a hand-based driving illegal behavior detection method, which comprises the following steps: reading images from a monitoring video; carrying out pretreatment on the read image, wherein the pretreatment comprises format change, establishment of an ellipse skin color model and determination of skin color probability values of pixel points; detecting and positioning a hand contour from the images obtained after pretreatment; with a steering wheel being the center, defining five regions of interest, and carrying out coordinate division on the defined regions; judging whether hand gestures are illegal driving behaviors by judging whether the hand contour locates in a center region of the steering wheel; carrying out hand pixel information preliminary determination by establishing the ellipse skin color model and utilizing logistic regression analysis; carrying out connected region analysis and labeling through a region growing method, and extracting hand skin color images and obtaining a complete hand image; and dividing the image into five regions of interest, so that whether operation in a driving cab is illegal can be judged accurately.

The present invention relates to an adipocyte automatic counting method based on image segmentation and edge detection. According to the method, threshold segmentation and edge detection are performed on an adipose image; the grayscale difference of an object and a background is extracted; the image is divided into a target area and a background area which have different grayscale levels; the edge of the image is a set of pixels of which the image grayscale goes through spatial sudden change; edge pixels in the image are determined, and then are connected together to form a border; adipocyte edges are extracted; morphological processing and filtering operation are performed on the adipocyte edges, segmentation is performed through a watershed algorithm, and therefore, under-dyed adipocyte edges can be identified and segmented, and therefore, the accuracy of adipocyte counting can be improved; connected region analysis is performed on the image, so that the number, area and other data of adipocytes can be put into statistics, and therefore, the efficiency of the counting of the adipocytes can be greatly improved; and dying processing and labeling are performed on the image, and therefore, subsequent manual correction can be facilitated, and high counting accuracy and efficiency can be ensured.

The obtaining condition input interface receives a gesture input while a subject appearance image (human body icon) is displayed, the obtaining condition identification unit identifies, based on a type of the gesture determined by the gesture type analysis unit and a region of the subject corresponding to the gesture identified by the gesture region analysis unit, a medical information obtaining condition for obtaining medical information corresponding to the gesture, the medical information obtaining unit selectively obtains medical information satisfying the identified medical information obtaining condition from the medical information database storing a plurality of sets of medical information, and the medical information display control unit displays the obtained medical information on the display means.

The invention provides an identification and analysis system for jewelry jade and rock minerals. The identification and analysis system comprises a detection system, a spectrum analysis system, a database and a result display system, wherein jewelry jade to be detected is detected by the detection system so as to obtain detection data, then the detection data flows to the spectrum analysis system, comparison is carried out by the spectrum analysis system to obtain a result, and then the result is transmitted to the result display system; the detection system is an infrared spectrometer; the spectrum analysis system comprises 400-1500 cm<-1> fingerprint region analysis, 150-4000 cm<-1> group frequency region analysis and near-infrared region 4000-7500 cm<-1> analysis. The system provided by the invention has the advantages that testing wave bands and testing methods (transmission method, reflection method and compression method) are expanded in the transverse direction, a data spectrogram base is established in the longitudinal direction, a set of system is formed from the detection and analysis of samples and presentation of result, the detection can be efficiently completed in a time-saving and labor-saving manner, and the accuracy of jewelry jade is improved; meanwhile, new information of jewelry jade can be added at any time.

The invention discloses an infrared and visible remote sensing image registration method based on salient region analysis, which can be used for registration of infrared and visible remote sensing images in aerospace and aviation. The method comprises the steps of: acquiring infrared and visible images of the same scale and direction according to longitude and latitude information; extracting salient regions beneficial for feature matching in the infrared images according to the feature consistency principle; extracting SIFT (Scale Invariant Feature Transform) feature points in two images and calculating feature descriptors; and carrying out feature matching according to the similarity, the scale and the distance constraint of the features; carrying out weighting and sequencing of the feature matching results according to salient images; selecting a subset of matching results with translation consistency; calculating the transformation matrix of maximal inner points of the subset according to the selected feature points; and transforming the infrared images into the same coordinate system as the visible images. Based on the method, the influence of wrong matching results on results is restrained, and the problem of infrared and visible image registration caused by different sensor types is solved.

The invention discloses face detection method and system for detecting driver fatigue. The method comprises the following steps of: image preprocessing: greying a color image and reducing the image resolution; image processing: acquiring a connected region and an integral image, wherein the acquisition of the connected region is realized by carrying out image binaryzation, edge image extraction, face region decomposition and connected region analysis on the current frame image, and the acquisition of the integral image is realized by carrying out image difference on every two adjacent frame images; selecting the face region of the candidate according to the acquired connected region and integral image; and verifying the face region of the candidate, filtering out the required false face region by using the judgment conditions, and outputting the face region. The invention can be used for detecting driver fatigue, realizes the driver's face detection, and has the advantage of simple and quick detection.

This invention presents an apparatus and system and its accompanying method for automatically placing of RFID antennas and a method for obtaining a readable region for passive RFID Tags are disclosed. The apparatus and system include a Portable and Adjustable Mounting Framework (PAMF) with built in chain belt for holding the RFID antennas, an Antenna Position Controller (APC) and Antenna Position Holders (APH) for control the movement of the antennas, a Moveable Object Platform (MOP) for carrying a number of RFID tags and a Computer Integrated System (CIS) with Readable Region Analysis (RRA) software to capture the transmitted and received RF signal from / to the RFID readers.

The invention provides a method for dispatching and allocating shared automobiles based on hotspot region analysis. The method comprises the steps of firstly, acquiring vehicle location related information uploaded by all shared automobiles when being used within a predetermined time period; secondly, dividing an overall region in which the shared automobiles are utilized into a plurality of sub-regions; thirdly, working out hotspot regions in the plurality of sub-regions according to the acquired vehicle location related information; and fourthly, enabling the number of the shared automobiles allocated to the hotspot regions to be greater than that that of the shared automobiles allocated to non-hotspot regions when the shared automobiles are dispatched and allocated.

The invention discloses an insect dimension measurement method based on Rayleigh region analysis scattering modeling, and the method comprises the steps: carrying out the Rayleigh region scattering modeling of an insect through employing a rotational symmetry dielectric ellipsoid; measuring the three-dimensional axon direction, body length and width of the insect through at least two radars, thereby achieving the further inversion of the mass and the body length to body width ratio of the insect. Compared with a conventional insect radar dimension inversion method, the method can greatly increase the amount of obtained information, so the method can improve the insect type recognition capability of the insect radars.

The invention discloses a finite-stage machine-based method for detecting an abandoned object. The finite-stage machine-based method comprises the steps of: firstly, establishing a Gaussian mixture background model, respectively establishing a short-time background model and a long-time background model according to different learning rates; secondly, establishing a finite-stage machine for each pixel; with results obtained by detection of two different background models with different updating speeds as inputs and a pixel classifying result output by each finite-stage machine as one binary image, carrying out communicated region analysis on the obtained binary image to obtain the shape and contour of the abandoned object by using a region growth method, and working out a rectangular box enclosing the abandoned object; and finally, timing the obtained rectangular box of the abandoned object, and detecting the abandoned object and alarming when a threshold is reached. According to the finite-stage machine-based method, under the condition that the abandoned object is absorbed by two backgrounds, the abandoned object can be always detected according to the pixel history, less false-alarms and wrong-alarms exist in an actual scene, and a good detection effect can also be obtained in an occasion with intensive pedestrian flow without depending on tracking information.

The invention provides a method and equipment for detecting the deviant behavior of a database. The method comprises the following steps: collecting SQL (Structured Query Language) sentences corresponding to the normal behavior of the database; carrying out SQL sentence keyword analysis, SQL built-in function analysis, SQL expansion storage process analysis and SQL sentence sensitive region analysis of the SQL sentences; generating a white list database; and adopting a formula V=f1*w1+f2*w2+f3*w3+f4*w4 to obtain an analysis result V of the database, and if the analysis result V is larger than or equal to a preset value, determining the behavior of the database is deviant. The method and the equipment for detecting the deviant behavior of the database, which are provided by the invention, can comprehensively detect the database, avoid omission and improve the efficiency of detecting the deviant behavior of the database.

The invention discloses a method for judging traffic illegal behaviors of a key vehicle not running in a regulation line. The method comprises the steps of: generating an objective key vehicle passage record set through a highway checking point device; sieving road elements corresponding to the reported route from a highway network map layer based on the reported route of the key vehicle, generating an initial planned route GIS map layer, and then obtaining crossed points to generate the GIS map layer of the points must be passed in the initial planned route; performing map exhibition on the passage record set by utilizing the GIS system, establishing a path automatic calculation process after buffer region analysis and ambiguous point removal, generating the practical route track GIS map layer of the vehicle, and generating a new planned route GIS map layer by combining the GIS map layer of the points must be passed in the initial planned route with a spatial topological relation of the road network; overlaying the planned route GIS map layer and the practical route track GIS map layer in different colors on a road network data set, and judging the route matching degree automatically by a system. According to the invention, whether the vehicle runs in a regulation line is judged through the running track formed by checking point passage information, and installation of a GPS (Global Positioning system) device is unnecessary.

The invention discloses a poor quality area analysis method, apparatus and system, and relates to the technical field of communication. The method comprises the following steps: analyzing key performance indicator KPI data to obtain network indicator data used for evaluating network quality; evaluating network data within grids representing different areas in a communication system by using the obtained etwork indicator data to obtain a plurality of poor quality grids with poor network quality; performing clustering processing on the plurality of poor quality grids to obtain a poor quality area with poor network quality; and performing analysis processing on associated ticket data and related service data in the poor quality area to find the reason for the poor network quality. By adoptionof the poor quality area analysis method, apparatus and system provided by the invention, the system capacity is effectively improved, the processing efficiency of the system is improved, and the time is shortened.

The invention discloses an address analysis method and a device based on a word segmentation algorithm. The method comprises the following steps: collecting the administrative division data of the National Bureau of Statistics and storing the data in a database; encoding the region name; establishing a region cascade relationship; setting up a region cascade relationship; setting up a region cascade relationship; constructing the regional decision tree, taking the country as the root node of the regional decision tree, province / municipality as its lower node, and creating the sub-node and leafnode of the regional decision tree recursively according to the regional cascade relationship; according to the full name and abbreviation of the region, the user-defined region dictionary being built and loaded into the database; acquiring regional information, segmenting the regional information with a word segmentation algorithm and a user-defined regional dictionary to obtain a regional phrase; according to the order of regional phrases after word segmentation and regional decision tree, address information being obtained. The invention solves the problems of low resolution accuracy and long time consumption of the short text region analysis function provided in the current map tool.

The invention discloses a compound cleaning method of a rubber product mould. The method comprises the steps that S1, region analysis is conducted, wherein before cleaning, dirt accumulation conditions of the mould before and after utilization are compared, and region division is conducted on the mould according to the compared conditions; S2, a sodium bicarbonate or potassium bicarbonate solution with the mass fraction being 10-20% is sprayed on the surface of the mould until the surface of the mould is saturated, then 3-25% alums are evenly sprayed on the surface of the mould, and the mould is subjected to standing for 3-10 minutes; S3, laser cleaning is conducted; S4, after laser cleaning, water is sprayed on the surface of the mould to conduct cleaning. The method has the advantages of high rubber product mould cleaning speed, low consumed preparation amount, little water pollution and low operation cost.