57results about How to "Avoid misidentification" patented technology

The invention provides a fatigue driving monitoring method and system and a storage medium, and the method comprises the steps: obtaining a DMS video stream, carrying out the feature extraction of the DMS video stream, and obtaining a face ROI region; inputting the face ROI region into a regression model to extract skin brightness change characteristics of the driver, and obtaining multiple vital signs of the driver based on the skin brightness change characteristics; the vital signs at least comprise heart rate, respiratory rate, oxyhemoglobin saturation, blood pressure and heart rate variability; the vital signs are judged according to a preset time period and multiple vital sign threshold ranges, and when at least two of the multiple vital signs exceed the corresponding vital sign threshold ranges, it is judged that the driver is in the fatigue driving state. According to the method provided by the invention, multiple vital signs are detected by depending on one independent regression model, and the accuracy and timeliness of fatigue driving detection are improved.

The invention provides a self-supervised mulberry disease and insect pest recognition method and system fused with a multi-mode YOLO algorithm, and the method comprises the steps: providing a Multi-Modan-YOLO algorithm fused with a multi-mode information processing technology, namely, the multi-mode YOLO algorithm, and carrying out the self-supervised mulberry disease and insect pest recognition method fused with the multi-mode YOLO algorithm; and audio information and thermal imaging information are combined to further improve the accuracy of the identification result of the pest and disease damage types. A self-supervised learning framework driven by a multi-modal large model is introduced, when the identification deviation risk exceeds a preset threshold value, the multi-modal large model is called to generate a category label, and the model is endowed with the ability of dynamically generating a new label for a new type of pest and disease damage. Therefore, misrecognition caused by non-uniform distribution of training data or missing of a certain category of data is effectively prevented; according to the method, targets such as disease spots and insect pests on mulberry leaves are accurately identified, so that the efficiency and accuracy of mulberry disease and insect pest prevention and control are improved, the influence on silkworm production is reduced, and yield increase is promoted.

The utility model provides track inspection equipment. The track inspection equipment comprises a vehicle body, and a steel rail detection device, a fastener detection device, an evacuation platform detection device, a ballast bed detection device, a drainage ditch detection device, a synchronous positioning system and a main control unit which are arranged on the vehicle body, wherein the steel rail detection device, the fastener detection device, the evacuation platform detection device, the ballast bed detection device and the drainage ditch detection device are all in communication connection with the main control unit, and the main control unit is in communication connection with the synchronous positioning system. According to the utility model, steel rails, fasteners, evacuation platforms, ballast beds and drainage ditches around the vehicle body are detected according to image information, the detection content is more, the accuracy is high, the daily public work detection content of urban rail transit is basically covered, the labor intensity is effectively reduced, and the detection efficiency is high.

The invention discloses a fingerprint identification method and system based on an optical identification induction technology, and relates to the technical field of fingerprint identification, and the method comprises the steps: collecting a fingerprint surface texture image through the total reflection principle of light; converting the collected fingerprint surface texture image into a fingerprint surface texture grayscale image; determining the main direction and frequency of the fingerprint streaklines, and separating the adhered fingerprint streaklines; smoothing stray point information in the fingerprint image by using an image filtering algorithm; the fluctuation characteristics of the fingerprints are further highlighted by using morphological expansion and corrosion operation; extracting morphological features of fingerprints and local features of streaklines from the preprocessed fingerprint image; and outputting the user information corresponding to the fingerprints in the database higher than the preset similarity value. By arranging the image acquisition module, the image processing module, the feature extraction module and the similarity comparison module, the interference of environmental factors is reduced, the fingerprint image quality is ensured, the fingerprint identification time is shortened, and the real-time performance of the system is improved.

The invention relates to the technical field of electric power communication module meter reading test, and discloses an electric power communication module meter reading test method and system, and the method comprises the steps: sending a to-be-tested module into a conveyor belt, carrying out the light supplement after the to-be-tested module enters a detection region, collecting an image, and recognizing display data. If the data display is abnormal, further judging whether bright spots exist or not, and if not, removing abnormal modules; and if yes, adjusting the light supplementing strategy and re-collecting the image until the bright spots disappear. And comparing and synthesizing the newly collected image and the original image, and determining whether the ammeter is qualified after determining that the data is accurate. Light supplementing adjustment is based on analysis of the bright spot position, the light supplementing effect is optimized by controlling the position of a light supplementing lamp, and the accuracy of detection data is ensured. The method can effectively improve meter reading accuracy and reduce manual intervention, and is suitable for automatic production line detection of intelligent electric meters.

The invention discloses a word document-based knowledge graph generation method and apparatus, and an electronic device, and relates to the field of data processing. The method comprises the following steps of: analyzing a target document to obtain a corresponding xml (Extensible Markup Language) file; the xml file is analyzed to obtain a first data structure, and the first data structure comprises a plurality of elements; in the first data structure, sorting the plurality of elements according to a preset sorting rule to obtain a second data structure; segmenting the second data structure by adopting a preset recursive function to obtain a plurality of entities and a relationship among the plurality of entities; and taking the plurality of entities as nodes of a preset knowledge graph, and taking the relationship among the plurality of entities as edges of the preset knowledge graph to obtain a target knowledge graph. According to the technical scheme provided by the invention, the document content is converted into the XML format for recognition and extraction, so that the accuracy of converting the document into the knowledge graph is improved.

The present disclosure relates to an autonomous mobile device, a control method and apparatus thereof, and a storage medium, the method comprising: a shooting step of shooting an image of a working environment of the autonomous mobile device; a distance measurement step: obtaining point cloud data of a working environment of the autonomous mobile device; an identification step: inputting the image into an image object identification model for identification to obtain an object in the image and a space range of a detection frame where the object is located; an obtaining step: obtaining the number of the point cloud data in the space range of the detection frame; a judgment step: judging whether the quantity of the point cloud data exceeds a preset quantity threshold value or not; and an object positioning step of calculating the spatial position of the detection frame according to the point cloud data in the spatial range of the detection frame under the condition that the number of the point cloud data exceeds a preset number threshold. Therefore, wrong object recognition or wrong object position judgment can be avoided, the accuracy of obstacle recognition and obstacle avoidance of the autonomous mobile equipment is improved, and thus the intelligent degree of the autonomous mobile equipment is improved.

The invention provides a coal mine rock burst prevention method and device, electronic equipment and a storage medium. The method comprises the following steps: acquiring first monitoring data of a coal seam roof rock stratum, wherein the first monitoring data at least comprises a fracture event and micro-fracture distribution density of the coal seam roof rock stratum; determining that the rock burst risk exists in the coal seam roof rock stratum in response to the fracture event and the micro-fracture distribution density meeting the risk identification condition; at least one time of pressure relief treatment is conducted on the coal seam roof rock stratum, and different pressure relief agents are used in the at least one time of pressure relief treatment process; and second monitoring data of the coal seam roof rock stratum after pressure relief processing are obtained, and when the second monitoring data do not meet the risk identification condition, it is determined that rock burst risk prevention and control are completed. Therefore, according to the scheme, combined monitoring of various data can be achieved so that the rock burst risk can be accurately recognized, roof pressure relief is achieved through cooperative prevention and treatment of at least one time of pressure relief treatment, and therefore the pressure relief effect is improved.

The invention discloses an audio data processing method and device, a storage medium and electronic equipment. The method comprises the following steps: determining audio data required in a target model training process; determining a target spectrum based on the audio data; window division and splicing are carried out on the target spectrum to obtain a first feature sequence, the first feature sequence comprises a plurality of first features recorded according to a specified time sequence, and first overlapping data exist between two first features which are continuous in time; and determining audio features based on the first feature sequence. According to the method, window division and splicing are carried out on a target frequency spectrum to realize data enhancement of audio features, after data enhancement processing, long-sequence audio data are split into a plurality of data fragments with overlapped parts, multiple times of reasoning can be carried out, and the data fragments have relevance, so that the accuracy of reasoning is improved. Misrecognition caused by single reasoning of the target model can be effectively prevented, and the recognition accuracy of the target model is improved.

The invention relates to the technical field of energy storage power supply management, in particular to a control method, system and device for an energy storage power supply battery pack and a medium, and the control method for the energy storage power supply battery pack comprises the steps: obtaining a plurality of interrupt signals generated by insertion of each battery module in real time, and sorting the interrupt signals to obtain an interrupt signal sequence; closing a corresponding interrupt pin based on a signal sequence in the interrupt signal sequence, and reading a default ID of each battery module connected to the host; generating corresponding battery numbers according to the default IDs of the battery modules, integrating the battery numbers to form a battery number list, and storing the battery number list in a host; and obtaining each battery state based on the battery number, obtaining battery current data according to each battery state, and performing energy equalization processing on each battery pack based on the battery current data. The method has the effects of automatically identifying the battery packs and realizing energy balance among the battery packs.

The invention provides a watermark recognition method and device, electronic equipment and a readable storage medium. The method comprises the steps of obtaining a to-be-recognized image; and inputting the to-be-recognized image into a pre-trained watermark recognition model to enable the watermark recognition model to calculate the contribution probability of each image position in the to-be-recognized image to the watermark classification result, determining the output probability of each watermark type included in the to-be-recognized image, and outputting the watermark recognition result in the to-be-recognized image. Therefore, wrong recognition of the watermark in the complex to-be-recognized image can be avoided, and the watermark recognition accuracy of the watermark recognition model can be improved; meanwhile, a plurality of synthesized sample watermark images carrying automatically labeled watermark type labels and watermark mask labels are adopted to train the watermark classification model, so that the problems of training sample richness and few samples are solved, and the accuracy of the trained watermark recognition model on watermark recognition is further improved.

The invention discloses an unmanned aerial vehicle water-air dual-medium adaptive communication optimization method and related equipment. The method comprises the following steps: acquiring unmanned aerial vehicle environment sensing information; determining the current communication medium state of the unmanned aerial vehicle according to the unmanned aerial vehicle environment sensing information; the communication medium state comprises an air state, an underwater state and a water-air interface crossing state; generating a target communication strategy according to the current communication medium state of the unmanned aerial vehicle, and controlling the unmanned aerial vehicle to execute the target communication strategy; the communication strategies comprise an air communication strategy, an underwater communication strategy and a water-air dual-medium communication strategy; and dynamically monitoring the communication link performance index of the unmanned aerial vehicle, and dynamically optimizing the target communication strategy in combination with the environment change information and the communication link performance index. The technical problem that the communication performance of the unmanned aerial vehicle on the water-air interface is sharply reduced can be effectively solved, smooth transition and high-reliability maintenance of the communication link in the double-medium environment are achieved, and the method can be widely applied to the technical field of unmanned aerial vehicle communication.

The present application provides an instruction processing method, a controller, a backlight driver and a display device, wherein the controller generates a driving instruction, the driving instruction includes a reset instruction, a setting instruction, an address instruction and backlight brightness data arranged in sequence, and then a preset mark is added to the driving instruction, and the preset mark indicates that the backlight brightness data is written into a data register. Subsequently, the driving instruction is sent to the backlight driver, so that after the backlight driver recognizes the preset mark, it can only perform a write operation on the data register in the backlight driver to write the backlight brightness data into the data register, thereby preventing the backlight brightness data from being mistakenly recognized as a setting instruction or an address instruction, and reducing the occurrence of display abnormalities.

The present invention discloses a playing method and system of a smart speaker, and the playing method includes: S1: receiving a voice command obtained by a Bluetooth headset through the voice recognition function of the Bluetooth headset itself; S2: based on the voice command, sending a sound content acquisition request to a background server to obtain the sound content; S3: judging whether the sound content can be publicly played; if so, executing step S4; if not, executing step S5; S4: publicly playing the sound content, and synchronizing the sound content to the Bluetooth headset for playing; S5: sending the sound content to the Bluetooth headset for playing. The technical solution of the present invention can synchronize the sound content of the smart speaker to the Bluetooth headset for playing to meet the different needs of users; it can also recognize the user's voice command through the Bluetooth headset voice to solve the problems of limited voice recognition distance and low voice recognition accuracy when using smart speakers for voice recognition.

The present invention provides a sequential AI diagnostic model for a chest X-ray intelligent diagnosis system, including an image classification module that identifies an orthotopic chest X-ray image in a DICOM image; an image quality judgment module that extracts orthotopic chest X-ray images that meet preset conditions and defines them as qualified images; a postoperative and implant diagnosis module that detects whether there is a postoperative or implanted object in a qualified image; a lung field segmentation diagnosis module that segments the lung field on a qualified image without a postoperative or implanted object; a binary classification diagnosis module that performs a positive and abnormal diagnosis on the lung field; an AI solution management module that determines the running order of each module, and sends the qualified image and the diagnostic data output by the previously running module to the later running module according to a preset configuration; and a structured report module that automatically generates a diagnostic impression based on the diagnostic data of each module. The present invention also discloses a sequential AI diagnostic model for a chest X-ray intelligent diagnosis method. The present invention improves the accuracy of diagnosis and reduces the cost of inspection.

The invention belongs to the field of artificial intelligence, and relates to an automobile part identification method, which comprises the following steps: acquiring image data acquired by a mobile terminal camera, and acquiring multiple frames of to-be-identified images from the image data according to a preset time interval; performing component detection on each frame of the to-be-recognized image by adopting a pre-trained automobile component recognition model to obtain a component detection result corresponding to each frame of the to-be-recognized image; the component detection result corresponding to each frame of the to-be-recognized image is sequentially stored in a cache queue according to the timestamp of the to-be-recognized image, and the cache queue only reserves the component detection result of a preset duration; and outputting a component identification result of the image data according to the component detection result in the cache queue and a judgment result of a preset rule. In addition, the invention also relates to a block chain technology, and the recognition result of the automobile part and the like can be stored in a block chain. According to the invention, the accuracy of automobile part identification can be improved.

This invention provides an automatic fiber extraction method, a training method for a fiber segmentation model, and an electronic device. The automatic fiber extraction method includes: segmenting a patient's medical image using a fiber segmentation model to obtain a segmentation result; wherein the fiber segmentation model is a pre-trained deep learning model for extracting slender structures; and performing connected component analysis on the segmentation result to extract the fiber region. This invention pre-trains a fiber segmentation model, which is particularly suitable for extracting slender structures in three-dimensional medical images, improving the accuracy and efficiency of fiber segmentation. After obtaining the fiber mask through segmentation, connected component analysis is performed to further accurately screen out the fiber region, eliminating interference from artificial filling materials, blood vessels, and other noise, thus more accurately determining the fiber region.

The invention discloses an industrial robot image processing method based on image fusion, relates to the technical field of image fusion, effectively solves the problem of object occlusion in storage through multi-angle real-time image acquisition and a deep learning hierarchical fusion algorithm, and improves the image fusion accuracy by evaluating the visibility coefficient Zjy of an occlusion area. Identification errors caused by shielding are avoided; the image denoising coefficient Ncj is calculated in real time, so that the image quality is ensured, and the subsequent recognition accuracy is improved; for shape and color diversification, a color diversification characteristic coefficient Ssc and a shape complexity coefficient Xtx are calculated, and a color histogram and geometric morphology analysis are combined, so that the recognition capability of a complex object is comprehensively improved; through comparison with the preset threshold value, the diversification conformity of the object is judged, objects of various shapes and color combinations are effectively recognized, the data labeling requirement is reduced, and the recognition efficiency and accuracy of an automatic warehousing system are improved.

The application relates to the technical field of wet tissue production equipment, and the wet tissue detection device comprises a rack, the upper surface of the rack is fixedly connected with a conveying mechanism, the conveying mechanism can send a wet tissue body to be detected into a detection area along a specified direction, one side of the conveying mechanism is fixedly connected with a machine box, a liquid crystal touch screen is installed in the machine box, the upper surface of the conveying mechanism is fixedly connected with a light box for providing a light source, the surface of the light box is fixedly connected with a support, an industrial camera one is fixedly connected to the support, a material pushing air cylinder is fixedly connected to the inner wall of the support, and the upper surface of the conveying mechanism is provided with a posture adjusting device for adjusting the posture of the wet tissue body. The wet tissue detection device solves the problems that when the wet tissue passes through the detection area, part of the packaging corners exceeds the ROI detection range, the packaging defects in the exceeding area cannot be effectively identified, and finally the detection is missed, resulting in inaccurate overall detection result.

The application discloses a notebook shell identification stamping forming device and belongs to the technical field of notebook shell production and processing. The device comprises an upper die assembly, a lower die assembly, a first identification assembly, a second identification assembly and a rib member. The first identification assembly is arranged on the upper die assembly. The second identification assembly is arranged on the lower die assembly. The first identification assembly comprises one or more first identification blocks. The rib member is provided with one or more. The second identification assembly comprises one or more second identification blocks. One end of the first identification block protrudes from one side of the upper die assembly and the lower die assembly. The rib member is arranged on one side of the second identification assembly facing the upper die assembly, and the rib member is arranged in a circumferential direction along the edge of the second identification block and the first identification block. The rib member protrudes in the direction of the upper die assembly. The above technical scheme has the advantages of high processing efficiency and avoidance of the formation of an R corner at the top edge of the identification area on the shell.

The invention provides a flame morphology identification method, system and device in a heating furnace and a medium, and belongs to the technical field of image identification. The method comprises the following steps: acquiring multiple frames of images in a heating furnace in an actual combustion process; wherein the multiple frames of images are obtained by decompressing video information in the heating furnace in the actual combustion process; for each frame of image, identifying the frame of image by using different image identification algorithms to obtain the flame morphology of each frame of image under each image identification algorithm, and determining the flame morphology under each image identification algorithm based on the flame morphology of each frame of image under each image identification algorithm; and based on the characteristic parameters of the flame morphology under each image recognition algorithm, flame morphology correction is carried out, and the corrected flame morphology is obtained. According to the invention, the flame morphology in the heating furnace can be effectively identified.

This invention relates to the field of communication technology, specifically a communication data security analysis and processing method, system, and storage medium. The method includes: recording communication data for various IoT scenarios based on edge node monitoring; performing gradient aggregation on the edge node data and filtering target communication subsets for collaborative operation of edge nodes; extracting abnormal communication data under time-delay collaboration based on the logical coordination and temporal continuity of the target communication subsets, combined with the indirect access path composed of edge nodes and regional nodes; transforming the scenario label, node location, and time period of the abnormal communication data into constraints based on the abnormal time and abnormal response intensity associated with the abnormal communication data, and solving for risk collaboration factors when data is abnormal; monitoring scenarios where risk collaboration factors occur, and optimizing the blocking of indirect access paths when new risk collaboration factors are detected. This improves the accuracy and efficiency of communication data risk analysis.

The invention discloses a credible identification method and system for identities of admission personnel of a large-scale activity, and particularly relates to the technical field of identity identification, and the method comprises the steps: obtaining a first image and a second image of the mth target personnel, and carrying out the histogram statistics of the second image, and obtaining a second gray histogram; analyzing the second grey level histogram by using an automatic threshold algorithm, determining the window width and the window level of the second grey level histogram, and adjusting each grey level value in the second grey level histogram to obtain an adjusted thermal imaging image; according to the method, the thermal imaging image is adjusted through histogram statistics and an automatic threshold algorithm, the quality of the thermal imaging image can be optimized, blood flow distribution information can be enhanced, and then the recognition stability and accuracy, especially the performance under different illumination and environment conditions, are improved. According to the method, the advantages of visible light and thermal imaging are combined, so that the error-tolerant rate of the system to the identified object is higher, and interference factors such as illumination, shielding and makeup can be effectively resisted.

The present invention relates to the field of image processing technology, and in particular to a method for monitoring the quality of new printed materials based on computer vision. The method acquires an image of a new printed material; acquires the degree of speckle in the closed area according to the edge roughness and position distribution of the closed area in the image; acquires the background possibility degree according to the edge roughness and speckle degree of the closed area in each suspected background closed area; acquires the clustering distance of any two closed areas according to the difference in the background possibility degree and speckle degree of the suspected background closed area in which any two closed areas are located, and then clusters the closed areas to obtain cluster clusters; detects speckle defects according to the edge roughness of the closed area in the cluster cluster. The present invention accurately divides the closed areas in the image by acquiring the clustering distance, avoids the interference of normal areas on speckle defects, and is conducive to accurately identifying the speckle defect area.

The invention relates to the technical field of license plate recognition, and discloses a license plate recognition credibility judgment method based on multi-algorithm fusion, and the method comprises the following steps: S1, carrying out the image collection and recognition of a vehicle parked at a roadside parking space, and obtaining the license plate information and the vehicle information; s2, scoring and assigning the recognition result of the license plate according to the recognition condition of the vehicle; s3, the distance between the license plate and the camera is scored and assigned; and S4, adding the two scores, and judging whether the identified license plate result needs to be manually rechecked or is credible. A monocular distance measurement algorithm and a vehicle identification algorithm are introduced, a scoring system is added for an identified license plate result, and manual rechecking is introduced when the license plate score is lower than a threshold value; a non-license plate pattern is prevented from being mistakenly recognized as a license plate under the conditions of image distortion, insufficient light and complex shooting scene; therefore, data support is provided for a roadside parking space parking charging application scene, and generation of wrong charging of a parking lot is reduced.

A payment terminal comprising: an IC card reader slot in front of a user; and a first plate having a touch area for contactless payment, the first plate being arranged obliquely upward toward a user with an open space arranged between the IC card reader slot and the first plate and being provided outside a housing of the payment terminal, the first plate having a touch area for contactless payment, the first plate being arranged obliquely upward toward the user, the IC card reader slot and the first plate are arranged in this order from a plane in a height direction, an antenna for contactless payment is arranged in the first plate, and the touch area is an area touched by a user in a contactless payment operation. And the open space is arranged across a surface having the IC card reader slot and two side surfaces adjacent to the surface having the IC card reader slot.

The invention relates to the technical field of image recognition, and provides a license plate recognition method based on video stream deep learning, and the method is characterized in that the method comprises the steps: obtaining video stream data of a monitoring camera, carrying out the frame extraction processing, and obtaining a frame sequence image; preprocessing the frame sequence image, and detecting a license plate area by using a license plate detection model; performing character recognition on the detected license plate area image to obtain a single-frame license plate recognition result; and carrying out preferential selection on the plurality of single-frame license plate recognition results corresponding to the frame sequence image to obtain a final license plate recognition result. According to the method, a complete technical framework integrating image enhancement, deep learning detection and recognition and multi-frame result probabilistic voting preferential selection is constructed, so that real-time, high-accuracy and high-robustness automatic recognition and statistical analysis of license plate information in a video stream are realized.

The invention relates to the technical field of vehicles, in particular to an intelligent driving safety auxiliary method and device, a vehicle and a storage medium, and the method comprises the steps: obtaining an actual tracking period of a monocular vision sensor for tracking a front target; if the actual tracking period is smaller than the preset period, the detection distance between the vehicle and the front target and the first detection speed of the front target are determined according to images collected by the monocular vision sensor, and the second detection speed of the front target detected by other sensors is obtained; and fusing the first detection speed and the second detection speed to obtain a fusion speed of the front target, and controlling the vehicle to execute a safety auxiliary action according to the detection distance and / or the fusion speed. Therefore, the problems of low driving safety, poor driving experience and the like caused by poor target perception fusion adaptivity, possible false recognition, false braking of the vehicle due to low recognition precision of the driver and other targets in auxiliary driving in related technologies are solved.

The present invention discloses an open set finger vein recognition method and system based on contrastive learning and generative learning. The finger vein training image is input into the contrastive learning model for feature extraction, and the contrastive learning model is optimized based on the obtained contrastive learning features; the training image is input into the generative model to calculate the reconstructed image, and the generative model is optimized based on the reconstruction loss and KL divergence; the alignment loss is calculated based on the contrastive learning features and the generative learning features, and the generative model is further optimized based on the alignment loss and the reconstruction loss; the finger vein test image is input into the optimized generative model, and it is judged whether the reconstruction loss is greater than the preset loss threshold. If so, the test image is judged as an unknown category, otherwise, the test image is judged as a known category; the test image judged as a known category is input into the optimized contrastive learning model for category prediction. The distinction between known category images and unknown category images is achieved through the reconstruction loss threshold, and the erroneous recognition of unknown users is eliminated.

The present application discloses a target detection method, device, electronic device and readable storage medium, which is applied to the field of artificial intelligence technology. The method includes: obtaining an image to be detected, performing target detection on a target object in the image to be detected, and obtaining a target position feature corresponding to the target object; obtaining the similarity between the target position feature and at least one position reference feature, wherein the position reference feature is determined based on the position of the target reference object in multiple sample images; and obtaining a detection result based on the similarity and the corresponding similarity threshold. The target detection method proposed in the present application can improve the accuracy of target detection to a certain extent by matching the target position feature with at least one position reference feature.