291results about How to "Narrow detection range" patented technology

The invention provides a vehicle license plate recognition method, and relates to a method for image recognition. The method comprises the following steps of preprocessing an image; partitioning a vehicle region according to color and texture characteristics; extracting a remarkable factor graph of a vehicle region diagram; extracting candidate vehicle license plates by an Adaboost classifier based on expanded Harr-like characteristic; determining the position of a real vehicle license plate from the candidate vehicle license plates; partitioning the marked vehicle license plate from the corresponding vehicle region original diagram; carrying out character segmentation according to structural characteristic; and carrying out character recognition based on the improved template matching method. With the method provided by the invention, the defects that the application scene of the traditional vehicle license plate recognition method is relatively single, some traditional vehicle license plate recognition methods are only suitably used for single vehicle license plate recognition of the single scene and difficultly used for multiple-vehicle license plate recognition of multiple scenes, and the recognition rate is easily affected by strong light, haze and weak light environments are overcome.

The invention relates to a rapid detection identification method for a specified pedestrian or vehicle in a video monitoring network. According to a target template provided by a user, accurate and rapid detection identification is carried out on a specified target in a monitoring network video. In the method, firstly, normalized scale change is carried out on a target template image and a plurality of kinds of mixed characteristics of the template image are calculated; a mixed Gaussian model is adopted to carry out background modeling on each monitoring video, and then parallel processing is carried out and area filtering and morphological postprocessing are used to extract motion foregrounds; normalized scale change is carried out on all motion foregrounds which meet a preliminary screening condition and a plurality of mixed characteristics are calculated and then weighted similarity distances between the motion foregrounds and the target template are calculated and first N objects, which are smaller than a judgment threshold and have smallest weighted similarity distances with the target template image, are returned and used as detection identification results. On the basis of ensuring the detection identification accuracy, the method significantly improves the processing speed of an algorithm through methods of parallel processing and frame skipping and the like.

The embodiment of the invention discloses a method and apparatus of detecting image quality. The method of detecting image quality includes the steps: acquiring an image to be detected, and the face area information which is obtained by pre-performing face detection on the image to be detected and is used for indicating the face area in the image to be detected; based on the face area information,extracting the face image from the image to be detected; inputting the face image into a pre-trained convolutional neural network to obtain the probability that each pixel point of the face image belongs to the type indicated by each type identifier in a preset type identifier set; inputting the face image into a pre-trained face key point positioning model to obtain the coordinates of each facekey point of the face image; based on the probability and the coordinates, determining the probability that the face image is shielded; and based on the probability of being shielded, determining whether the quality of the face image is qualified. The method of detecting image quality improves the image quality detection efficiency.

In a hand drying apparatus, an infrared light absorber absorbs infrared light, when a hand is not present, emitted from an infrared light emitter provided on an inner surface of a drying space.

The invention discloses a video image processing-based fire detection system and detection method. A video acquisition module, a fire detection module, a fire analysis module, a fire judgment module and a fire alarm module are connected with one another so as to form the video image processing-based fire detection system; the video acquisition module acquires video images in the coverage area of a camera; the fire detection module performs background modeling and background updating on the video image and detects a foreground moving object; the fire analysis module analyzes the video images to identify a suspected flame region and extracts static characteristics and dynamic characteristics of the suspected flames; the fire judgment module adopts the static characteristics and dynamic characteristics of the suspected flames extracted by the fire analysis module as the input of a BP neural network so as to judge whether flames exist in a video; and the fire alarm module judges whether to give an alarm according to the output result of the fire judgment module. With an existing monitoring system used in combination, the video image processing-based fire detection system can decrease detection cost and greatly reduce a false alarm rate with fire detection accuracy ensured.

A fully automatic and energy efficient lighting control or light switch comprises different settings, preset by a user, and activates all or a portion of at least one bank of lights upon detecting occupants within a room. In one embodiment, the light switch detects doppler-shifted, reflected ultrasonic waves caused by occupant motion within the room. In an automatic mode, the light switch is configured to automatically switch from an initial sensitivity level for detecting motion within a short range therefrom, to a higher sensitivity level for detecting motion anywhere within the room. After the room is empty or no motion is sensed, the light switch is configured to turn the lights off following a variable time delay. The light switch returns to its initial low sensitivity level following a predetermined grace period which begins once the lights are turned off. The lights can be turned off manually when the light switch is in its automatic mode, in which case, the light switch continues to operate at the higher sensitivity level. When motion is no longer detected, and following the variable time delay and predetermined grace period, the switch automatically returns to an automatic "on" state. In an alternative manual mode, the light switch can be activated manually and deactivated both manually and automatically. In its manual mode, the light switch is configured to deactivate the lights upon sensing no motion and is configured to reactivate the lights automatically within the predetermined grace period upon sensing motion.

The invention discloses a monocular vision and laser radar fusion-based road travelable region detection method and belongs to the intelligent transportation field. Existing unmanned vehicle road detection methods are mainly based on methods such as monocular vision, stereo vision, laser sensor and multi-sensor fusion methods, have defects of low robustness to illumination, complex three-dimensional matching, laser sparseness, low overall fusion efficiency and the like. Although some supervised methods have achieved better accuracy, the training processes of the supervised methods are complex, and the generalization effects of the supervised methods are poor. According to the monocular vision and laser radar fusion-based road travelable region detection method provided by the present invention, ultra-pixel and point cloud data fusion is adopted; on the basis of features, road regions can be obtained through machine self learning; and the features are fused through the Bayesian frame, so that road information is obtained, and a final region can be obtained. With the method adopted, strong hypothesis information and complex training processes are not required. The monocular vision and laser radar fusion-based road travelable region detection method has the advantages of excellent generalization performance, high robustness, fast speed and high precision, and can be popularized and used more easily in practical application.

The invention discloses a cross-site script vulnerability detection method based on a percolation test. According to the cross-site script vulnerability detection method, the technical problem of low detection efficiency in a traditional cross-site script vulnerability detection method is solved. The technical scheme provided by the invention is that the cross-site script vulnerability detection method comprises the following steps of: firstly, acquiring a page; secondly, extracting form information and carrying out the percolation test on the extracted form; and finally, carrying out session retaining and redirection tracking. Due to the adoption of a session retaining and redirection tracking method, the conditions that the analysis for Web pages is incomprehensive are reduced, all pages of a Web site are analyzed and detected and further the automatic detection capability is improved. According to a percolation test strategy, the detection range of the XSS (Cross-Site Script) is gradually narrowed and the depth detection for the safe pages is avoided. While test statements are injected, test statements for closing HTML (Hypertext Markup Language) statements are actively constructed according to the returned page information, so that interaction frequencies with a server are reduced and the detection efficiency of the XXS vulnerabilities of large complex Websites is improved.

The invention discloses a face key point detecting method and device. The method comprises the steps that a face image to be detected is obtained; initial key points of N specific organs are determined in the face image by using a first face key point detecting algorithm, wherein N is larger than or equal to 1; the region, where each specific organ in the N specific organs is located, in the face image is determined with the corresponding initial key point as reference; a final key point of each specific organ in the N specific organs is determined in the region, where the specific region is located, in the face image by using a second face key point detecting algorithm. By means of the method, the number of the searching regions detected through the second face key point detecting algorithm can be reduced, and the detecting efficiency of face key points can be improved obviously.

The invention relates to a street litter recognition method. The street litter recognition method applied in a complex environment comprises the following steps that 1, a street picture is acquired, litter areas and litter-free areas are chopped, and a sample set is constructed for training a deep convolutional neural network (DCNN); 2, registration and pixel level differentiation are conducted on a to-be-detected real-time street picture and a clean street picture to obtain changed areas of the image, whether the areas are litter or not is judged according to the output vector of the deep convolutional neural network (DCNN), and if yes, marks are made on the real-time image. According to the method, detection errors caused by interference factors such as the complex environment and illumination changes can be effectively inhibited while the litter targets are not missed, and all-weather street litter recognition in the complex environment is achieved.

The invention discloses a fire video detection and early warning method based on image multi-feature fusion, and the method comprises the steps: firstly carrying out the preprocessing after an image sequence of a video is obtained; extracting a foreground region, and obtaining a detected candidate region; secondly, extracting static features and dynamic features from the candidate areas, judging whether flames are contained or not by taking the static features and dynamic features as input of an SVM classifier when the flames are detected, and obtaining whether smoke is contained or not afterlogic combination selection and calculation are conducted on feature judgment results when the smoke is detected; finally, if it is detected that flames or smoke exists, carrying out fire judgment according to the growth trend of the flames or smoke; when it is judged that a fire occurs, carrying out fire alarm on the monitoring site, and otherwise carrying out fire alarm only on the background. The method can be combined with an existing monitoring system to be applied to places such as shopping malls and warehouses, the fire detection and early warning cost is reduced, the detection method is good in generalization ability and applicability, reliable fire detection and early warning functions can be provided, and the method has practical value.

The invention provides an illegal service vehicle detecting method and system. Based on traffic information collected by a bayonet detection system, the vehicle activity degree in a road network is analyzed through a clustering statistical method. A legal service taxi operating mode is tapped. Illegal service vehicles are detected from a group with the highest degree of activity. According to the method and the system, automatic illegal service vehicle identifying, blacklist managing, investigating and handling and other functions are realized; automatic detection of vehicles in the road network is realized; illegal service vehicles are identified; and an offsite enforcement system is provided with a reliable basis, which is conductive to urban road traffic management.

The invention provides a blind decoding method for reducing blind decoding times of PDCCHs (Physical Downlink Control Channels) based on polar codes and belongs to the field of communication. According to the method, before polarization coding is carried out, maximum values of two DCI (Downlink Control Information) lengths under a current transmission mode are calculated and unified, moreover, two possible DCI formats are distinguished through setting of a one-bit flag bit, and CRC check bits are added behind the information with the flag bit, thereby obtaining an information sequence participating in the polarization coding. Codon sequences transmitted by the PDCCHs are obtained through punching polarization coding according to a PDCCH aggregation level required by a high-level signal. In a blind decoding process of a user side, an information bit length output through decoding is calculated according to a transmission mode, a PDCCH sequence consistent with a current user identity is screened through a CASCL decoding algorithm, the final DCI information length is determined according to flag bits in a decoding result, and control information is extracted. According to the blind decoding method provided by the invention, on the premise of not improving an original decoding error rate, the decoding times is effectively reduced, the system efficiency is improved, and the practical application is facilitated.

The invention discloses a battery appearance defect detection method based on dimension reduction and point cloud data matching, belonging to the technical field of machine vision detection, which obtains three-dimensional point cloud data of a battery to be detected and reduces the dimension of the point cloud data. Obtaining a defect area of the battery to be detected, and extracting point clouddata of the defect area; Extracting the point cloud data of the same area of the standard battery; Sampling two pieces of point cloud data and matching searching; acquiring The optimized correspondence of feature points and calculating roughly rigid transformation relation to realize rough registration. Performing Constraint detection on the rough registration results, and verifying correctness.Optimizing the rigid transformation relationship between the point cloud data to achieve automatic and accurate registration to determine whether the battery appearance is qualified or not. Changing A3D imagen into a 2D image by a dimension reduction algorithm, and acquiring a point cloud data of a defect area by using a plane defect detection technology in that 2D image so that the point cloud data is matched, the detection range is narrowed, the running time is reduce, and the accuracy is improved.

The invention relates to the defect detection technology field and discloses a defect detection method based on visual identification and a system. The method comprises steps that an image of a to-be-detected panel is acquired; edge detection on the image is carried out, and a target contour is extracted; the target contour is a contour on which defect detection is carried out; a first tangential gray scale gradient is calculated; the first tangential gray scale gradient is a tangential gray scale gradient of a first pixel point on the target contour; the first tangential gray scale gradient is equal to a difference value of a gray scale of a first pixel and a gray scale of a second pixel; the second pixel is adjacent to the first pixel and is before the first pixel on the target contour; according to the first tangential gray scale gradient, whether defect is detected is determined; if the tangential gray scale gradient satisfies a preset defect determination condition, that defect is detected is determined. Through the method, automatic detection on fine defects in a panel printing process is realized, a detection result is objective, and dependence on artificial experience can be avoided.

The utility model relates to a detection method of eye state based on cascade sort and Hough circle transform, belonging to the field of pattern recognition, which is characterized in the following steps: to acquire a face image; to carry out skin color segmentation in the YCbCr color space to acquire the position information of the skin color area using an ellipse skin model; to detect the rectangular eye area with the method of eye detecting window traversal with a cascade eye classifier; to merge a rectangle by rectangular merge method to get a merged eye rectangular linked list; to operate edge detection and binarization in turn on every rectangular eye area in the merged eye rectangular linked list with a Sobel operator to get an binary image; to detect the eye state of every binary image in turn with horizontal projection method; to further detect the eye state by Hough circle transform detection method if whether the present state is eye closure is not certain. The utility model increases the speed of eye detection, tracks and analyzes the eye state with the skin color segmentation, which is suitable for attention detection and fatigue detection.

The invention relates to a correction method for an incomplete or deformed quadrangular image. The method comprises the five steps of edge point detection, abnormal edge point elimination, linear fitting, vertex calculation and image correction. The correction method for the incomplete or deformed quadrangular image is applied to an identification method identification system for certificates such as valuable documents, notes, identification cards and the like, and can be used for rapidly performing edge detection and tilt correction on defect and deformed valuable documents, notes or certificate images.

The invention proposes a multi-feature fusion cascaded classifier-based fire fighting access vehicle detection method, and belongs to the field of image processing and machine learning. The method comprises the steps of extracting a road region in a fire fighting access image by utilizing a significance detection method so as to eliminate the influence of an object, in a non-lane region, on a detection result; extracting a vehicle bottom shadow region by adopting a vehicle bottom shadow adaptive threshold segmentation method based on a characteristic that a grayscale value of the road region obeys normal distribution; obtaining a lower edge of the vehicle bottom shadow by adopting a pixel change rate-based shadow edge extraction method, and constructing a region of interest of a vehicle according to the extracted lower edge of the vehicle bottom shadow; and extracting an Haar feature of the region of interest of the vehicle, training a feature value after fusion and an HOG feature value of ''integrogram'' calculation by utilizing an Adaboost algorithm to generate a strong classifier, and finally verifying whether the region of interest of the vehicle is the vehicle or not by using a cascaded classifier obtained by training.

In a hand drying apparatus, an infrared light absorber absorbs infrared light, when a hand is not present, emitted from an infrared light emitter provided on an inner surface of a drying space.

The invention discloses an SAR image change detection method based on SIFT feature points, and mainly aims to solve the problem that the conventional detection method is highly influenced by 'spot' noise. The SAR image change detection method comprises the following steps: (1) conducting normalization processing on two SAR images which are rectified and are subjected to radiation correction and geometric correction; (2) acquiring a difference image according to the two normalized images to be detected; (3) extracting SIFT feature points in the difference image by using a scale invariant feature transform method; (4) conducting region growing by taking the SIFT feature points as seed points, thereby obtaining a variation image. The method has the advantage of high 'spot' noise robustness and can be applied to SAR image variation detection under a low signal-to-noise ratio.

The invention provides a slant vehicle detection and tracking system and method based on machine vision. The system comprises a CCD camera, a USB data transmission port and a computer terminal. The method comprises the following steps of: in vehicle detection phase, firstly, image preprocessing is performed, region-of-interest extraction and improved lane line detection are combined, an inclined vehicle detection area is divided; an adaptive threshold value and a maximum between-class variance method are adopted; a shadow area is extracted, further a shadow line is extracted at the bottom of the vehicle, left and right boundaries of the vehicle are determined in combination with Sobel vertical edge extraction to obtain a suspected rectangular frame of a vehicle target, then features in the rectangular frame are extracted, dimensionality reduction is performed on the features by adopting kernel principal component analysis, and detection confirmation is performed by utilizing an Adaboost cascade classifier; in the vehicle tracking stage, mean shift and Kalman filtering are combined, a vehicle detection result is used as an initial tracking target, and the tracking target is screened by using rectangular frame coincidence. The system and the method can realize real-time vehicle detection and tracking, and have high accuracy.

The invention discloses a beam recognition method, a beam recognition system and a network node. The method beam recognition comprises the steps that: a first network node sends one or a group of beam recognition auxiliary information to a second network node, wherein the beam recognition auxiliary information corresponds to one or a group of high-frequency sites and is used for indicating beam information required for accessing the high-frequency site to the second network node; and the second network node obtains the one or a group of beam recognition auxiliary information, and recognizes an optimal beam of the high-frequency site according to the obtained beam recognition auxiliary information. By adopting the beam recognition method, the beam recognition system and the network node, since the second network node obtains the beam recognition auxiliary information required in the beam recognition process, the detection range in the access process is reduced; in addition, when compared with the method of accessing the high-frequency site through blind test, the complexity of optimal beam recognition is reduced, the time delay in the optimal beam recognition process is shortened, and further the access rate to the high-frequency site is increased.

The invention discloses a quick positioning method of facial feature points. The quick positioning method comprises the following steps: 1) taking Haar-Like features as a basis, and adopting a trained cascaded classifier to successively detect a left eye and a right eye; 2) taking a detection result of the left eye and the right eye as the basis, obtaining the precise positions of the inner canthus point of the left eye and the inner canthus point of the right eye; 3) taking the inner canthus points and the distance between the inner canthus points as a criterion, applying a geometric feature of "Three Tings and Five Eyes" of the human eyes to quickly determine key feature points on the outer contour of a mouth and the outer contour of each eye, storing the points as a reference feature table to serve as a search reference of a next frame of feature points; and 4) taking a previous frame of a left-eye coordinate stored in the reference feature table as an initial position, taking inner canthus distance as the criterion, setting a search range of a current frame of human eyes at a ratio according to a geometry of a human face, repeating the steps 1) to 3), carrying out full-figure search again when the left eye can not be detected within a set search range, and repeating the steps 1) to 3). The quick positioning method can quickly and accurately position the facial feature points without a special device.

The invention provides a curving lane line detection method and device and a terminal, which relate to the technical field of auxiliary driving. The method comprises the following steps that the edgebinary image of a road image to be detected is longitudinally segmented into a preset number of sub-images; in the edge binary image, a road vanishing point and a lane line of the sub-image are determined according to a candidate lane line detected in the sub-image; whether a bent part exists in the lane line or not is determined according to the discrete degree of the road vanishing point and theslope of the lane line; if the bent part exists, the edge binaryzation is carried out and a curve which is used for representing the bending part of the lane line is fitted in the image. By applyingthe method, the curved lane line can be accurately detected, so that a lane deviation early warning system can be assisted to realize correct early warning as far as possible.

The embodiment of the invention provides a memory leak detecting method and a memory leak detecting device. A memory comprises a doubly linked list of an allocated memory; and the doubly linked list comprises memory header information of one or more allocated memory blocks. The method comprises the steps of: determining a detection time interval; traversing the header information in the detection time interval in the doubly linked list; and in accordance with the memory header information, determining whether the corresponding one or more allocated memory blocks are subject to memory leak. By the memory leak detecting method and the memory leak detecting device, memory leak can be visually detected in real time, the software performance is improved, and manual operation is reduced or avoided, so that the method and the device are labor-saving and time-saving, the operation difficulty is reduced, the working efficiency is improved, the cost is reduced, the accuracy of the detection of memory leak is improved, and the leaked memory can be recycled in the entity life cycle.

The invention discloses a flow analysis-based IOS malicious software early warning and detection system and method, and relates to the field of mobile terminal information security. The system comprises an IOS mobile phone detection unit (10) and a server detection unit (20); the IOS mobile phone detection unit (10) and the server detection unit (20) interact for realizing functions of detecting and analyzing malicious software in an IOS; and an analysis report is formed and provided for a terminal for early warning. The method comprises (1) an IOS mobile phone malicious software pre-judgment method, and (2) a server virus research and judgment method. The system and the method have the following advantages and positive effects: (1) monitoring of a non-jailbreak version is supported originally; (2) the detection range is wide, and the malicious software and bundled software are supported; (3) a detection scheme is rigorous; (4) system expandability is high; and (5) traceability is achieved.

The invention relates to a phosphorescent energy transfer system, a synthetic method and use of the system and a detection method of thrombin. The detection method comprises the following steps: fixing an aptamer with an amino-modified nucleic acid at 5'-end on the surface of an Mn doped ZnS phosphorescent quantum dot wrapped by mercaptopropionic acid (MPA) through dehydration and condensation reaction, wherein the marked phosphorescent quantum dot is close to the surface of a carbon nano point and phosphorescence is quenched; and by means of bond tight effect of the nucleic acid aptamer and the thrombin, separating the phosphorescent group from the surface of the carbon nano point, wherein the phosphorescent strength is recovered for measuring the amount of the thrombin. In the experiment, the content of the thrombin in a living body is detected by the phosphorescent energy transfer principle between the phosphorescent quantum dot and the carbon nano point, so that the detection selectivity and specificity are improved, the phosphorescent background is reduced, the quenching time is greatly shortened, the detection sensitivity is improved, the limit of detection is further low, and the phosphorescent signal collected is stable, thus the demand of microanalysis is met.

This disclosure provides an underwater detection device, which includes a first generator for generating a first transmission signal, a second generator for generating a second transmission signal having a longer pulse width than that of the first transmission signal, an output unit for outputting the transmission signals generated by the first and second generators to a transducer for transmitting underwater ultrasonic waves based on the first and second transmission signals, respectively, receiving the reflection waves caused by the ultrasonic waves, respectively, and converting the reflection waves into a first reception signal and a second reception signal, respectively, an underwater detector for detecting at least one of a presence of a target object and a water bottom depth based on the first reception signal, and a bottom sediment discriminator for discriminating a water bottom sediment type based on the second reception signal.

The invention provides a method for detecting a car safety belt fastening state based on a road monitoring device. The method comprises the steps of: 1, obtaining a monitor image from the road monitoring device; 2, positioning a car body area in the original monitor image; 3, positioning a car window area in the car body area by hough conversion; 4, positioning the position of a steering wheel in the car window area based on arc detection; 5, extracting the image of a safety belt area according to the position of the steering wheel; 6, linearly detecting the safety belt, detecting that whether a straight slope corresponding to the safety belt is within 0.75-1.2, and thereby judging whether a driver in the car window area wears the safety belt; according to the method, the area where the safety belt is located is positioned by using the car body and car window areas and the steering wheel position, the detection range is gradually reduced, accuracy and efficiency in safety belt detection are improved, automatic judgment can be performed instead of using human eyes, and thereby accuracy and efficiency in detection on fastening of the safety belt are improved.