113results about How to "Reduce false recognition rate" patented technology

The invention provides an access control system based on face recognition and a recognition method, and the system comprises a multi-modal image collection subsystem which integrates a visible light camera, a depth camera and an infrared thermal imaging camera and is used for synchronously collecting a visible light image, a depth image and an infrared thermal imaging image of a face; the image preprocessing unit is used for carrying out corresponding preprocessing on the image; the feature extraction network module is used for extracting image features; the multi-modal feature fusion and alignment module is used for calculating a mutual information matrix by adopting a feature fusion strategy based on mutual information maximization and adaptive weighting, determining an adaptive weight and performing weighted summation, and performing feature alignment by applying a method based on canonical correlation analysis; and the face recognition classification and verification module verifies the image by adopting a multi-level fusion strategy. According to the method, the false recognition rate and the rejection rate are effectively reduced, the strong anti-cheating capability is achieved, and cheating attacks of various fake face means can be accurately recognized and resisted.

The invention relates to the technical field of voice recognition, in particular to a mixed recognition processing method and device, equipment and a medium, and the method comprises the steps: recognizing command word data based on a command word recognition model, selecting command word data with an error rate lower than a preset range, generating a corresponding phoneme embedding representation, and forming a phoneme embedding dictionary; determining a mixed recognition command word corresponding to the current command word, and generating a mixed recognition list; identifying an audio embedding sequence corresponding to the current command word; obtaining a phoneme embedding sequence of each mixed recognition command word in the mixed recognition list based on a phoneme embedding dictionary; and calculating the similarity between an audio embedding sequence and each phoneme embedding sequence to obtain an identification result. The method and the device can improve recognition accuracy and reduce mixed recognition of command words.

The invention discloses a method and device for recognizing and analyzing PDF texts and pictures based on OCR and a storage medium, and particularly relates to the field of text and picture processing.The method for recognizing and analyzing PDF texts and pictures based on OCR comprises the following steps that S100, PDF documents are loaded, page analysis is conducted, page elements are extracted, and page layout information is processed; and step S200, for the extracted picture or the PDF page generated by scanning, performing high-precision OCR recognition by using a deep learning model, and converting characters in the image into editable texts. According to the method, efficient and accurate analysis of complex texts and picture contents in the PDF document is realized, the technical scheme not only improves the analysis precision and efficiency and enhances the flexibility and expandability of the system, but also promotes the development of digital and intelligent application of the document, and meanwhile, by saving resources and time cost, the method and the system have good application prospects. And remarkable economic benefits are brought to enterprises and individual users.

The invention relates to the field of image recognition, and discloses an industrial drawing key symbol semantic recognition method and system, and the method comprises the steps: designing a general drawing segmentation model DrawSAM based on a basic model SAM, carrying out the self-adaptive fine adjustment of the SAM through employing a high-efficiency fine adjustment method DoRA, carrying out the image segmentation through combining with a Box Generator prompt generator, and carrying out the recognition of key symbols of an industrial drawing. The method can adapt to diversified drawing types and complex structure layouts, and the detection and recognition accuracy and efficiency of characters and symbols in the drawings are improved; and through the optimized text detection algorithm DrawDETR, the error recognition rate during engineering drawing recognition is reduced, the recognition accuracy is improved, the problems of error recognition and incapability of symbol recognition of a general OCR model in the field are effectively solved, retrieval and application of drawing electronic data are remarkably improved, and industrial requirements are met.

The invention belongs to the technical field of natural language processing, and discloses a natural language processing platform and method based on deep learning, and the platform comprises a data collection module which is used for collecting font data, text data, image data and environment data; the data processing module is used for preprocessing the collected font data, text data, image data and environment data and obtaining a comprehensive feature data set; the character recognition module is used for predicting character recognition results through a character recognition model and judging whether the character recognition results are consistent or not; the semantic recognition module is used for generating corresponding character paraphrases through a semantic recognition model; collecting character recognition results and corresponding character paraphrases through the software terminal; the voice output module is used for converting the character recognition result and the paraphrases of the corresponding characters into natural voices and outputting the natural voices by adopting a TTS technology; through the implementation of the technology, students can be helped to better learn and understand.

The invention relates to the technical field of electrical fire identification, in particular to an electrical fire intelligent identification method and system based on a visual image, and the system comprises a visual image fire identification platform, a fire information base, an infrared image analysis unit, a region identification unit, a feature matching unit, a dynamic management and control unit and an alarm response unit. According to the method, regional fire monitoring feedback analysis is performed on the infrared feature image preliminarily, so that the suspicious risk condition of each monitoring point in the target electrical equipment can be known, and data support is provided for subsequent analysis; whether the target electrical equipment has a fire or not is analyzed from the two aspects of infrared feature image processing and monitoring video image feature extraction, and through the mode of combining the two fire analysis methods, the fire recognition error rate is reduced, the target electrical equipment with a disaster is analyzed deeply, and the fire recognition efficiency is improved. Reasonable management and control can be carried out according to the dynamic change trend of the fire, and the fire management efficiency can be improved.

The invention relates to the technical field of identity verification, in particular to an identity verification method and system based on multi-modal data fusion, and the method comprises the following steps: based on user login information, comparing the login information with pre-stored information, recognizing a user identity, and extracting pre-stored face data and iris data of a user to obtain a user identity information extraction result. According to the method, the accuracy of identity verification is improved by combining face and iris recognition technologies, optimizing image quality and comparing pre-stored data, the possibility of wrong recognition and identity fraud can be remarkably reduced, and flexible response can be performed according to different security requirements by dynamically adjusting a matching degree verification threshold value; according to the method, the adaptability and the protection degree of the verification process are enhanced, the false identification rate and the false rejection rate can be effectively reduced, meanwhile, a stricter security verification process is provided, higher-level personal identity protection is realized, identity embezzlement and other related fraudulent behaviors are effectively prevented, and the trust of consumers to digital services is enhanced.

The invention relates to the technical field of gate channel target image recognition, and particularly discloses a gate channel target recognition detection method and system, and the method comprises the steps: collecting image data, carrying out the image preprocessing steps, such as illumination compensation and noise removal, extracting the brightness features and texture features in an image, and calculating the visibility coefficient of a target region. According to the method, high identifiability of the target is kept under various illumination conditions, the final identification accuracy of the target in different environments is evaluated by comprehensively analyzing the visibility coefficient and the identification confidence of the target, so that the image preprocessing and target identification algorithms are dynamically adjusted, the target can be rapidly and accurately identified, and the identification accuracy of the target is improved. The target identification efficiency in the gate channel is improved, the high efficiency and robustness of the system under the complex illumination condition are ensured, and the method has practical value and application prospect.

The invention relates to the field of campus energy-saving management, and discloses a smart campus energy-saving management method and system based on the Internet of Things, and the method comprises the following steps: transmitting collected power utilization data to a cloud server for storage; ultrasonic signals of the campus building are obtained and transmitted to the cloud server; defects existing in the campus building are recognized, and defect repairing is carried out; formulating energy-saving strategies of classes and public places; if the daily power consumption of the class or the public place exceeds the daily energy-saving power consumption target power, feeding back to a manager; the system comprises a power utilization data acquisition module, an ultrasonic data acquisition module, a model construction module, a power utilization equipment adjustment module and an energy-saving feedback module. The defects of the building structure can be found and repaired in time, potential safety hazards can be prevented, the thermal insulation performance of the campus building can be guaranteed by repairing the campus building, the place where water possibly leaks can be found in time, and energy consumption of the campus building is reduced.

The invention discloses a multi-mode defect automatic classification method, device and equipment and a storage medium. The method comprises the following steps: acquiring a microscope image, an X-ray image and defect feature information data of a to-be-detected product; first classification processing is preferentially carried out based on the microscope image, when the first classification processing result meets a first preset reliability condition, the first classification processing result is taken as a final classification result, and when the first classification processing result does not meet the first preset reliability condition, the final classification result is taken as a final classification result. Performing feature fusion based on the X-ray image and the defect feature information data, and executing second classification processing; when the second classification processing result meets a second preset reliability condition, the second classification processing result serves as a final classification result, and when the second classification processing result does not meet the second preset reliability condition, defect type judgment is conducted based on key parameters in the defect feature information data, and the final classification result is output; the classification accuracy and efficiency can be improved, and the error recognition rate is reduced.

The invention discloses a fluorescence spectrum analysis method for reducing an algae misrecognition rate based on spectrum similarity, belongs to the technical field of marine water ecological environment monitoring, and can effectively reduce the algae spectrum misrecognition rate. Compared with a standard method, the classification measurement result of 95 pure algae samples by the algorithm provided by the invention has four difference value points outside a 95% consistency limit, and the proportion is less than 5%. Compared with the national standard method, the classified measurement results of 49 algae samples in the mixed sample have three difference points outside the 95% consistency limit, and the proportion is also less than 5%. Considering the randomness of the measurement error, the result shows that the measurement result of the algorithm and the measurement result of the national standard method have good consistency.

The invention provides a human body behavior recognition method based on long-time-sequence motion semantic coding, and relates to the technical field of video analysis, and the method comprises the steps: S1, inputting a video image, detecting a pedestrian target in the image, and calculating the intersection-union ratio of pedestrian detection frames in a front frame image and a rear frame image to track the human body target; s2, extracting skeleton key points of each human body target by adopting a posture estimation algorithm, and defining the skeleton key points of the human body targets as middle-layer motion semantic features; s3, establishing a 3D CNN network model based on the middle-layer motion semantic features to extract short-time-sequence motion semantic codes; s4, constructing a long-time-sequence motion semantic coding model according to the short-time-sequence motion semantic coding so as to output a one-dimensional feature vector; s5, constructing a human body behavior classifier based on a full-connection network to recognize the one-dimensional feature vector, and outputting a recognition result; the accuracy of behavior recognition can be improved, and the problems that the error recognition rate is high, and application of a real scene is difficult to meet are effectively solved.

The invention relates to the technical field of classroom intelligent terminals, in particular to a classroom intelligent terminal integrating a micro host, network access and audio acquisition. According to the technical scheme, the classroom intelligent terminal integrating a micro host, network access and audio acquisition is provided. Comprising a main control unit, an acoustic array device, a network interface device, a deep learning audio processing unit, a multi-mode interaction device, an adaptive network technology component, an edge computing unit, a virtual reality and augmented reality device, a block chain module and a voice synthesis and emotion recognition module. The system provides more advanced, intelligent and interactive rich education experience, is expected to improve the learning effect and teaching efficiency, has potential revolutionary influence in the education field, and provides a more flexible, personalized and efficient learning environment.

A glass surface stain image recognition method applied to a window-cleaning robot relates to the technical field of image recognition, and comprises the following steps: acquiring image data, environmental perception data and tactile data of a target glass area according to a sensor terminal of the window-cleaning robot; constructing a shooting parameter lookup table, and updating shooting parameters of the sensor terminal in real time according to the environment sensing data in the current acquisition period and the shooting parameter lookup table; constructing a stain feature recognition model, and performing pre-estimated stain type labeling on the image data of the current acquisition period according to the stain feature recognition model; and performing correctness detection on the estimated stain type marked in the target glass area according to the tactile data of the current acquisition period, if a wrong stain type exists in the target glass area, obtaining the recognition correct rate of the stain feature recognition model, and performing local feature enhanced recognition operation or retraining operation according to the recognition correct rate. And the stain identification capability of the window cleaning robot is obviously improved.

The invention provides a biological multi-modal feature recognition method and device based on dynamic fuzzy reasoning. The method comprises the steps that an obtained face image, a fingerprint image and an iris image of a target object are preprocessed; performing feature extraction on the preprocessed face image, the fingerprint image and the iris image, and determining a three-dimensional Euclidean distance of the face, the fingerprint and the iris between the target object and an object to be verified in a database according to a feature extraction result; performing fuzzy reasoning based on the three-dimensional Euclidean distance of the face, the fingerprint and the iris to obtain a recognition score; comparing the identification score with a preset decision threshold value, when the identification score is greater than or equal to the decision threshold value, determining that the target object is matched with the to-be-verified object, otherwise, determining that the target object is not matched with the to-be-verified object; wherein the decision threshold value is obtained by training through a reward / punishment mechanism based on preset target recognition precision. The method can effectively cope with various environment changes, improves the recognition accuracy, and is suitable for application scenes needing extremely high safety.

The invention discloses a small sample face recognition method based on prototype calibration and adaptive fine tuning, which comprises the following steps: 1) pre-training an encoder on a large-scale face database, capturing rich semantic information and class distribution, constructing an evaluation set, and ensuring that training identity information is not overlapped with the constructed evaluation set; 2) initializing the weight of the classifier through a weight imprinting method, ensuring smooth transition, and inducing a discrimination structure of the encoder; 3) calculating the cosine similarity between the fine-tuning prototype and the pre-training prototype, fusing the pre-training information through weighted average, and calibrating the fine-tuning prototype with deviation; 4) introducing an adaptive interval loss function, adjusting the penalty intensity of the sample, and performing selective fine adjustment on the BatchNorm layer and the classifier weight; 5) using the domain perception similarity NAC to calibrate unknown category rejection, enhancing the similarity relation with the domain sample, and obtaining the final matching score. The method can improve the recognition precision and rejection capability of the model in a few-sample learning scene.

The present application relates to the technical field of sample pretreatment automation, and discloses a system of a full-automatic sampling tube cap screwing and pipetting workstation and a running method thereof.The system comprises a motion point definition module, a cap screwing adaptation identification module, a label scanning module, a pipetting planning and execution module, a cap screwing torque calling module and a real-time torque acquisition module.The system realizes continuous scanning and reliable identification of the cap profile in the motion state through a dynamic coordinate system and key motion points, dynamically calls matched torque parameters to execute cap screwing according to the identification result, and simultaneously acquires real-time torque data.The system synchronously completes sample label information analysis, pipetting path planning and pressure monitoring.The scheme realizes adaptive perception and closed-loop cap screwing control of the cap state, and improves the robustness of cap identification, the consistency of cap screwing action and the reliability of the overall process.

The embodiment of the application relates to the computer technical field, and provides an audio recognition method, related device and storage medium, the audio recognition method comprises the following steps: obtaining audio to be processed; performing frequency domain conversion on the audio to be processed to obtain target frequency domain acoustic characteristics of the audio to be processed, and performing feature extraction on the audio to be processed to obtain target audio attribute characteristics of the audio to be processed; performing feature fusion on the target frequency domain acoustic characteristics and the target audio attribute characteristics to obtain target audio characteristics; performing recognition on the target audio characteristics to obtain a recognition result, and the recognition result is used for indicating whether the audio to be processed is fake. The embodiment of the application can fuse the target frequency domain acoustic characteristics and the target audio attribute characteristics to obtain target audio characteristics containing richer and more comprehensive audio representation, so that accurate recognition can be carried out based on the target audio characteristics, and the accuracy of fake trace detection on the audio is improved.

The present application relates to the technical field of face recognition, and particularly relates to a recognition method and system based on a dynamic face database, a computing device and a storage medium, comprising a face tracking server configured to acquire face information obtained by a channel camera and a regional camera, and to merge the face information obtained by the channel camera and the regional camera; a central face server configured to store face information left by a user when registering remotely, and to respectively identify face information collected by the face tracking server from the channel camera and the regional camera, and to send the identification results to a channel face server and a regional face server; the present application reduces the number of face databases by using face tracking and face recognition technology, forms a face server with multiple face databases, improves the recognition efficiency by accurately simplifying the face databases, and reduces the false recognition rate of face recognition.

The application discloses a portrait inspection and identification method based on gait cycle decomposition and multi-phase force analysis, and comprises the following steps: extracting the skeleton key points of a target object in a target object image set in a comparison video and the confidence score of each skeleton key point; extracting the features of the target object in the target object image set according to the skeleton key points of the target object, and obtaining the target object features, wherein the target object features comprise basic features, gait cycle features, gait stage features and gait parameter features; integrating the basic features, the gait cycle features and the gait stage features to obtain comprehensive features; respectively calculating the similarity of the target object features and the comprehensive features in the comparison video to obtain similarity calculation results and overall similarity calculation results; and obtaining the identity identification result of the target object according to the similarity calculation results and the overall similarity calculation results.

The invention discloses a device for automatically recognizing pictures, characters and sound in real time, which comprises a data acquisition module used for acquiring picture, character or sound information of a specified target; the extraction module is used for extracting knowledge points in the picture, character or sound information; the knowledge base module is used for storing knowledge data information; the comparison module is used for comparing the knowledge points extracted by the extraction module with the knowledge data information in the knowledge base module to obtain comparison information; and the correcting module is used for correcting the knowledge points of the specified target according to the comparison information. Compared with the prior art, characters, pictures and sounds can be automatically recognized, the purpose of automatically correcting homework is achieved, and therefore the correcting time of teachers can be shortened.

The invention discloses a handwritten formula correction and complementation system, relates to the technical field of artificial intelligence, and solves the problem that in the prior art, the recognition effect is greatly reduced due to poor handwriting. For a long-sequence complex formula, more characters are included, the characters are densely arranged during handwriting, and writing is small, so that the problem that an identification result is wrong or missed is solved. Whether the handwriting input data is valid or not is judged according to the handwriting input signal; judging whether the noise level of the preprocessed handwriting input track data is normal or not; a handwriting recognition model is called and initialized, and the segmentation processing result is recognized through the handwriting recognition model; judging whether the user needs to perform manual processing according to the confidence of the recognition result; an intelligent prediction algorithm is introduced, and a historical input mode of a user and a common mathematical formula structure are combined to dynamically adjust an identification result; whether correction and complementation of the handwritten formula are completed or not is judged according to the operation of the user, and the accuracy and continuity of formula recognition and correction are greatly enhanced.

The present invention discloses a manuscript image recognition system and method based on fusion detection, belonging to the technical field of optical character recognition. The method specifically comprises: acquiring a manuscript image, and preprocessing the acquired manuscript image, performing angle recognition on the preprocessed manuscript image, and performing rotation correction processing, recognizing the text position in the manuscript image after the rotation correction, and recognizing the text content in the manuscript image based on the recognized text position in the manuscript image. By fusing a plurality of detection methods, the advantages of each method can be combined, the defects of a single method can be reduced, and the recognition accuracy of the manuscript image can be improved. The method is applicable to various complex manuscript image recognition scenarios, including but not limited to manuscript images in the financial field, etc.

The invention discloses an API path aggregation identification method and device and a storage medium, and relates to the technical field of electronic digital data processing, and the method comprises the steps: carrying out the preliminary aggregation identification of API path data based on the path structure features of the API path data; if the preliminary aggregation identification is not successful, performing deep aggregation identification on the API path data based on word segmentation features and / or statistical features of the API path data; and generating an aggregation identification result of the API path data. According to the method and the device, the API path aggregation identification accuracy is improved.

The invention provides a self-service vending method, a vending device, a vending room and a storage medium, the method is applied to the vending room, the vending room is provided with a delivery channel, and the method comprises the following steps: obtaining activity information of a user entering the vending room; information of taking the commodity by the user is acquired, and first prompt information is pushed, so that the user places the commodity on the inner side of the delivery channel; identifying commodity information on the delivery channel; generating first order information; pushing the first order information and confirming the first order information by a user; acquiring payment information and opening a door; by setting the delivery channel, the user puts the commodities into the delivery channel after selecting the commodities and identifies and confirms the commodities, so that the intelligence of commodity settlement is improved, the commodity identification error rate caused by misoperation of the user is reduced, and the consumption experience of the user is improved.

The invention provides an animal identification method and device, equipment and a storage medium, and the method comprises the steps: carrying out the marking of the categories and specific names of a sample image containing an animal, making a data set according to the categories, and training the data set to obtain a first classification model, making a data set according to the specific names, and training the data set to obtain a second classification model, and detecting the family of the animal and the position of the animal in the image through the first classification model, determining a second classification model based on the family of the animal, and identifying the specific name of the animal through the second classification model. According to the method provided by the invention, the accuracy of inter-department animal identification and intra-department animal identification is improved, and compared with the prior art, the manufacturing cost of the data set is greatly reduced.

This application provides a training method for a fingerprint recognition network, a fingerprint recognition method, an apparatus, and a medium. The training method for the fingerprint recognition network includes: determining fingerprint feature matching maps based on a first initial fingerprint image and a second initial fingerprint image; forming a training sample set from multiple fingerprint feature matching maps; and training an original neural network based on the training sample set to obtain a fingerprint recognition network. The technical solution provided by this application solves the problem of difficulty in distinguishing whether two different fingerprint images belong to the same finger, thus reducing the false acceptance and false recognition rates of fingerprint recognition.

The invention relates to an unmanned aerial vehicle image high consequence area building identification method, system and device based on priori knowledge and a medium, and the method is characterized in that the method comprises the steps: obtaining a to-be-identified remote sensing image collected by an unmanned aerial vehicle at a set height and route parameters; inputting the obtained to-be-recognized remote sensing image into a pre-trained multi-scale space attention aggregation model to obtain a building probability graph of a high-consequence area building recognition result; wherein the multi-scale space attention aggregation model is constructed based on a U-Net model and is provided with a multi-scale space attention aggregation module and a self-adaptive loss function, and the multi-scale space attention aggregation module is used for carrying out multi-scale space attention aggregation processing on an input feature map; according to the method, the building in the remote sensing image acquired by the unmanned aerial vehicle can be accurately identified, and the method can be widely applied to the technical field of computer vision and remote sensing image processing.

The application relates to a multi-protocol adaptive control method, system, device and medium of an LED dimming power supply, which comprises the following steps: receiving a dimming signal output by an external dimming device, and preprocessing the dimming signal to obtain a target input signal; extracting protocol identification features from the target input signal, and constructing a protocol feature vector; matching the protocol feature vector with a preset protocol template library to generate a candidate protocol set and corresponding initial confidence; dynamically correcting the initial confidence by combining historical recognition results, signal quality indexes and output response consistency to determine a target protocol type; calling corresponding protocol analysis rules and control parameter templates according to the target protocol type to generate target dimming control parameters; and adjusting LED drive output based on the target dimming control parameters. The application has the effect of improving the adaptive identification capability of the LED dimming power supply for various dimming protocols.

The present invention provides a bone medical image damage detection method and system, and relates to the field of medical image processing technology. The method comprises: obtaining a bone binary image; obtaining the bone contour line in the bone binary image; determining the bone fracture judgment score; obtaining the midline of two adjacent bone contour lines; determining the bone crack judgment score; in the case where at least one of the bone crack judgment score and the bone fracture judgment score is greater than or equal to a preset first damage score threshold, setting a detection line parallel to the contour line in the background area outside the bone area; obtaining the bone fragment judgment score; determining the bone damage judgment score; if the bone damage judgment score is greater than or equal to a preset second damage score threshold, marking the bone medical image to be processed. According to the present invention, image processing technology is used to assist doctors in diagnosis, thereby improving the detection accuracy and efficiency of bone damage.