42results about How to "Solve misidentification" patented technology

The invention discloses a hierarchical multi-source data fusion method for a pipeline linkage monitoring network, which comprises the following steps: carrying out data level preprocessing for various primary linkage detection signals acquired by a sensor at a common node of the monitoring network by using wavelet transformation, and extracting leakage-sensitive characteristic parameters; establishing a characteristic level data fusion model based on an ant colony neural network, processing the leakage characteristic parameters extracted by various sensors on the node, and constructing an elementary probability assignment function of evidence according to the output result of the ant colony neural network; and carrying out evidence synthesis at a cluster-head node according to an evidence combination rule, and making final decisions according to a maximum trust value method. The invention provides the hierarchical multi-source linkage detection data fusion method from the data level and characteristic level to decision level, and solves the multi-source data processing problem of the pipeline linkage monitoring network; and the method utilizes the linkage detection information acquired by various sensors in the network so as to effectively improve the accuracy rate of leakage identification.

The invention discloses a complicated equipment acoustic fault recognition and location method. The method comprises steps of carrying out data pre?processing and fault feature extraction on vibration signals of a single sensor in a distributed equipment monitoring network, using a within-class between-class distance as an evaluation function of genetic algorithm and optimally selecting characteristic parameters sensitive to a fault, building single-value classifier?models based on support vector data description for initial fault recognition through training a normal class sample set, and constructing a basic probability?assignment?function of?an evidence according to output information of fault classifiers of different measurement points, adopting an evidence combination rule to combine an evidence set and making a decision result. The invention provides an equipment acoustic fault recognition and location method with vibration information hierarchically-converged, a recognition technology in a multi-feature comprehensive analysis mode is adopted, vibration information of different measurement points is made full use of, a strong generalization ability is provided, and the problem that precise mathematical?modeling is hard to be built for fault recognition and location due to complicated structure of the equipment can be solved.

The invention provides a method for automatically identifying the opening pin defect of a transmission line, which comprises the following steps: an image of the transmission line is obtained; then, the pixel coordinates of the insulator region in the image are obtained by using the trained shared convolution neural network to detect the transmission line map; at that pixel of the insulator region, a feature recognition algorithm is used to detect whether the connector split pin is defective or not according to the coordinates. The program can automatically identify the working condition patrol image, and automatically locate the defect position of the patrol pin in the image for the operator to judge, improve the patrol efficiency and ensure the safe and stable operation of the transmission line.

The invention discloses a parking payment system, which is characterized in that a vehicle owner terminal registers on a parking platform, logs in a platform account and opens an ETC payment functionof a specified license plate number; a toll collection system comprises a license plate recognition camera and a control server, the license plate recognition camera monitors the motion condition of avehicle on a lane, captures a vehicle head when finding that the vehicle is about to enter a parking area, acquires a vehicle image, performs image recognition on the vehicle and a license plate, generates a license plate number and sends the license plate number to the control server; the control server calculates the parking duration according to the vehicle parking time and generates a parkingfee payment order, and the control server sends the license plate number and the payment order to the parking platform for ETC payment; and the parking platform matches the license plate number withthe platform account, judges whether the platform account corresponding to the license plate number opens an ETC payment function, and judges whether the license plate number is in an ETC blacklist.

The invention discloses a skin recognition method based on saliency detection. The skin recognition method comprises performing face recognition and saliency detection on an image to be detected to obtain a face skin region, calculating the face skin region to obtain an average skin color and a skin color probability mapping table, and finally, performing skin color recognition on the image to be detected according to the skin color probability mapping table and obtaining the result map of the skin color probability of the image to be detected; the skin recognition method has the advantages that the collection of huge training data is avoided, the recognition result is not affected by the deviation of the training data, and as saliency detection is performed, non-skin regions in the face region can be excluded, and therefore, the accuracy of recognition is higher and the problem of misrecognition due to that the image to be detected is too bright or too dark is overcome.

The invention discloses a tennis ball positioning system and method based on visual identification. The system comprises an image extraction unit, a segmentation unit, an extraction unit and a positioning unit which are sequentially connected, wherein the image extraction unit is used for extracting image information in a field; the segmentation unit is used for receiving the image information andprocessing the image information; the extraction unit is used for extracting the outline of a tennis ball in a processed image, and the positioning unit is used for acquiring coordinates of the tennis ball; the segmentation unit comprises a filtering module, a conversion module, a searching module and a screening module. Based on a novel tennis ball identification and positioning principle, binocular vision is used for identifying the tennis ball and positioning the tennis ball, the problem of false identification caused by the adoption of laser radar and the ultrasonic technology is solved,and the hardware cost is reduced on the premise that the identification accuracy and the positioning precision are ensured.

The present invention relates to a method for cleaning a solar photosphere bright spot in an astronomical image and belongs to the fields of astronomical technique, image processing and data mining. The method comprises the steps of: firstly, identifying a bright spot by using a method that combines a Laplacian filter and morphological dilation and tracking the bright spot by using a method of three-dimensional communication; then, extracting a plurality of eigenvalues, with a low correlation. of the bright spot, and after the eigenvalue data is standardized, performing dimension reduction according to a contribution rate by using a principal component analysis method; and finally, cleaning the bright spot data subjected to dimension reduction by using a DBSCAN method so as to remove a non-bright spot structure. According to the method for cleaning the solar photosphere bright spot in the astronomical image provided by the present invention, it is unprecedented and effective to apply the DBSCAN method to cleaning the solar photosphere bright spot data, and the method largely solves the problem of error identification existing in a traditional identification method of solar photosphere bright spot. Meanwhile, more accurate bright spot data is cleaned for small-scale magnetic field researches and further researches on problems such as corona heating.

The invention relates to an operation maintenance instruction execution method based on human-machine natural language voice interaction, a storage medium and a mobile terminal and belongs to the technical field of internet. Because the method comprises a voice awakening process and a voice identification process, Chinese natural sentences obtained by identification are utilized and converted intopinyin character strings, the pinyin character strings are further matched with an operation maintenance manipulation regular expression set, and a corresponding operation maintenance instruction isexecuted according to a matched operation maintenance manipulation regular expression. Because the operation maintenance manipulation regular expression set comprises regular expressions which can bematched with general natural language sentences with operation maintenance instructions in a pinyin way, the related natural languages can be effectively identified, and the problem of machine identification errors caused by Chinese polyphonic characters is solved; the technical requirements for workers are reduced, the accuracy of identifying the operation maintenance instructions is improved, the operation mode of an operation maintenance system is effectively simplified, and the operation maintenance working efficiency is improved.

The invention relates to the field of automatic driving, and discloses an abnormal driving behavior identification method and system, and a vehicle comprising the system, the abnormal driving behavioridentification method comprises the following steps: obtaining vehicle driving behavior data, and determining whether the vehicle is in a suspicious abnormal driving behavior state based on the vehicle driving behavior data; if the vehicle is in the suspicious abnormal driving behavior state, obtaining current vehicle driving scene data; judging whether the suspicious abnormal driving behavior isan abnormal driving behavior or not according to the vehicle driving behavior data and the current vehicle driving scene data; according to the technical scheme of the invention, the current drivingscene information is introduced into the recognition process of the abnormal driving behavior of the vehicle, and the recognition accuracy of the abnormal driving behavior is improved.

The embodiment of the invention discloses a vein recognition method and a related product. The method comprises the steps of obtaining a first vein image of a first to-be-recognized part; performing first vein matching based on the first vein image to obtain a first matching value; determining a second to-be-identified part and an image acquisition strategy corresponding to the second to-be-identified part based on the first vein image and the first matching value; acquiring a second vein image of the second to-be-identified part based on the image acquisition strategy, and performing second vein matching based on the second vein image to obtain a second matching value; and determining a vein recognition result based on the first matching value and the second matching value. According to the embodiment of the invention, the vein recognition precision can be improved.

The invention discloses a scene recognition method, a scene recognition device and electronic equipment, relates to the technical field of information processing, and can solve the problem of low scene recognition accuracy of an existing scene recognition method. According to the invention, image scene identification is carried out by combining the scene identification algorithm, and the time information, the position information, the weather information, the temperature information and the like of the image, so that the problem of false identification caused by scene identification only basedon the algorithm can be avoided and thus the accuracy of image identification is improved.

The invention provides a currency identification method, a currency identification device and electronic equipment. A first identification model used for identifying currency types of banknotes and asecond identification model used for identifying authenticity of the banknotes are pre-trained; currency information of banknotes can be identified by processing images, manual operation is not neededin the whole process, and the method has the advantages of being rapid, efficient and low in cost. Besides, the second recognition model is added on the basis of the first recognition model, so thatthe authenticity of the banknote can be recognized, the problem that the model mistakenly recognizes a non-banknote object as the banknote can be avoided, and the recognition accuracy is improved.

The invention provides a gesture recognition method, a gesture recognition device, electronic equipment, a readable storage medium and a chip. The method comprises the steps: obtaining a hand region subimage in a target image, and determining the position information of a plurality of feature points corresponding to the plurality of feature points in the hand region subimage; determining a first position feature vector according to the position information of the plurality of feature points, the first position feature vector representing a relative position relationship of any one of the plurality of feature points relative to other feature points in the plurality of feature points; determining a second position feature vector according to the position information of the plurality of feature points, the second position feature vector representing an absolute position relationship of the plurality of feature points in the hand region sub-image; and outputting a recognition result of the hand region sub-image according to the first position feature vector and the second position feature vector. According to the invention, the gesture recognition stability is improved, the complex shooting environment can be dealt with, the user gesture can be accurately recognized under different visual angle conditions, and the user intention can be effectively captured.

The embodiment of the invention provides an agent detection method and device. The agent detection method comprises the following steps: acquiring a communication data packet with a target agent server; analyzing the communication data packet to obtain agent information of the target agent server; and sending an agent service connection request to the target agent server based on the proxy information, and detecting whether the agent information is valid information or not according to the connection request result. According to the embodiment of the invention, the agent detection reliabilityis improved.

A pedestrian identification method based on a light field camera and HOG and SIFT mixed features comprises the following steps: (1) using the light field camera to shoot a large number of pedestrian images as a positive sample of an experiment; (2) shooting a large number of non-pedestrian images by using a light field camera as a negative sample of an experiment; (3) obtaining an original 2D image and a depth map of field of the shot image by using Lytro Desktop software; (4) performing ROI region extraction on the original 2D image by using a method of combining HOG + SIFT mixed features andSVM, and judging whether pedestrians are included or not; and (5) identifying the depth map of the pedestrian judged in the step (4) by using the method of combining the HOG + SIFT mixed features with the SVM again, and judging whether the deceptive pedestrian is included or not. According to the invention, the characteristics that the light field camera can record the scene 4D information and acquire the depth image are utilized to eliminate false identification caused by pedestrian traffic, so that the pedestrian identification capability is enhanced, and the working cost is low.

The invention discloses an aviation cable coding visual identification method and device considering coding directivity, and the method comprises the steps: collecting a cable surface coding image, and carrying out the image preprocessing; acquiring upper and lower boundary points of the cable in the cable surface coding image, fitting a boundary and performing inclination correction, and correcting the cable to be in a horizontal direction; coding in the cable surface coding image is segmented and optimized to obtain a plurality of rectangular areas, and the plurality of rectangular areas are rotated to obtain a plurality of coding characters to be identified; inputting the plurality of to-be-recognized coded characters into a plurality of pre-trained classifiers used for respectively coding directions, and obtaining the coding directions and a coding recognition result through the plurality of classifiers. According to the method, the large classifier is divided into a plurality of small classifiers, so that classification categories are reduced, and the accuracy of code recognition is improved.

The invention relates to a rapid portrait recognition device and method for an indoor scene. The method comprises the steps of preliminarily recognizing an indoor image by adopting an image recognition model to obtain a portrait recognition result such as a human face, a gesture or the like; then carrying out static object elimination processing on the portrait recognition result; comparing the portrait recognition result with the contents stored in a static object storage region; when the eigenvalues and positions of the portrait recognition result and the contents stored in the static objectstorage region are the same and a writing time difference between the portrait recognition result and the contents stored in the static object storage region is larger than a preset time, indicatingthat the image recognition result is a static object, and not outputting the image recognition result; otherwise, indicating that the image recognition result is not the static object, and writing theimage recognition result into an output result storage region; and outputting the image recognition result through a recognition result output module. The problem of false recognition caused by the static object is effectively solved; the recognition accuracy is effectively improved; and the user experience is improved.