681 results about "Region detection" patented technology

The invention discloses a transformer substation defect identification method and system based on cooperation of large and small models. The method comprises the following steps: S1, acquiring image data and preprocessing the image data; s2, acquiring a scene understanding score, and if the score is higher than a preset scene threshold value, synchronously executing S3 and S4; otherwise, only executing S3; s3, inputting the image data into the small model to obtain a detection result, and if the detection result is higher than a first preset threshold value, directly taking the detection result as a current defect identification result; if the detection result is lower than the second preset threshold value, abandoning the detection result; if the image is between the first preset threshold value and the second preset threshold value, obtaining a region detection result through image cutting and local large model reasoning; s4, inputting the image data into the transformer substation defect detection large model to obtain a global detection result; and S5, obtaining a final defect identification result according to the small model and / or regional and / or global detection result. According to the invention, substation defect identification precision can be improved.

The invention discloses a natural resource supervision method based on three-dimensional GIS scene and video fusion, and relates to the technical field of intelligent natural supervision. Comprising the steps of collecting a video stream through a preset observation point, extracting a continuous frame sequence, and calling a space parameter set; performing dynamic region detection on the video frame sequence to generate change feature data; mapping the change characteristic data into a topological structure to be updated based on the projection matrix; calling a pre-constructed three-dimensional GIS scene model, and positioning an update area; according to the type identifier of the change feature data, executing an operation of deleting the associated geometry or constructing and fusing a new geometry; adding a state attribute label to the updated geometry, and storing an update model; the resource change rate is calculated, and when the change rate exceeds a preset industry threshold value, an early warning instruction is generated and pushed to the control terminal. And the space accuracy, the model updating efficiency and the early warning reliability of resource change identification in a complex natural scene are remarkably improved.

The invention discloses a multi-spectral image fusion building surface biological attachment area detection method and device and a medium, relates to the technical field of image processing, and discloses a multi-spectral image fusion building surface biological attachment area detection method comprising the following steps: based on a to-be-detected building, obtaining a visible light image collected by a visible light camera of an unmanned aerial vehicle, the multispectral camera acquires a multispectral image sequence based on at least two different frequency bands; registering the visible light image and the multispectral image based on a calibration and preprocessing module to obtain a pixel-level aligned visible light image and multispectral image sequence; and according to a deep learning feature identification module, identifying the visible light image and the multispectral image sequence after pixel-level alignment, and obtaining a pixel-level segmentation result of the organism attachment area of the to-be-detected building. Therefore, detection is carried out based on the unmanned aerial vehicle, deep learning and an image feature fusion mode are combined, and the building surface bioattachment recognition accuracy is improved.

The invention discloses a power line anomaly detection method and system based on an unmanned aerial vehicle. The method comprises the steps that a server analyzes a power line detection request to obtain a detection area and a detection time period; planning an initial detection track, generating a power line detection instruction based on the initial detection track, the detection area and the detection time period, and issuing the instruction to the unmanned aerial vehicle; the unmanned aerial vehicle executes an inspection task in the electronic fence, the initial detection track is dynamically corrected in the inspection process, and a flight time sequence image is collected; analyzing the flight time sequence image through a line anomaly detection model, and outputting a line anomaly detection result; an actual detection track is recorded, and abnormity marking is carried out on the actual detection track; and recording, encrypting and uploading an inspection log to the server. The method has the advantages that the detection efficiency, the intelligent level and the safety management capability of power line anomaly detection are greatly improved.

According to the method, firstly, SAR data are mapped to an optical image domain by using a conditional diffusion model, pseudo-optical images with consistent spatial spectrums are generated, and fusion distortion caused by difference of SAR imaging mechanisms in a traditional method is overcome; secondly, a refined cloud region detection mechanism of a Fmask cloud mask is introduced, a cloud pollution region and a cloudless region are dynamically distinguished in combination with an adversarial training strategy, and the problem that the cloudless region is mistakenly changed in the reconstruction process of an existing method is effectively solved; besides, a multi-source integrated sample data set is constructed in stages, and a training normal form of multi-index joint optimization of PSNR, SSIM and the like is adopted, so that texture detail recovery and spectrum fidelity of a thick cloud coverage area are realized in a complex scene. In downstream application tasks such as land utilization classification and disaster dynamic monitoring, the visual quality and the quantitative index of the cloud removal result have good effects, and reliable technical support is provided for high-precision reconstruction of remote sensing information of a multi-cloud area.

The invention provides a remote sensing building group depth detection and change intelligent evaluation method for urban planning, and the method comprises the steps: carrying out the multi-modal data collection of a target urban region, and carrying out the standardization processing, abnormal value elimination, multi-modal alignment and interference suppression processing, and obtaining a multi-modal data set; constructing a multi-branch hybrid network, and performing triplet feature extraction on the multi-modal data set through the multi-branch hybrid network to obtain a building group fusion feature map; performing change area detection and change type classification on the two building group fusion feature maps with different time phases to obtain building group change information; and constructing a dynamic evaluation index based on the urban planning core demand, and performing building group change evaluation in combination with the building group change information. According to the method, through multi-source and multi-modal data acquisition and processing, multi-branch hybrid network triplet feature extraction and change detection and dynamic evaluation index application, urban planning core requirements are adapted, and building group detection precision and change evaluation effectiveness are improved.

The invention relates to the technical field of bill text recognition, and discloses a bill text recognition system and method based on deep learning. The method comprises the following steps: acquiring target bill original image data containing a multi-channel pixel matrix and spatial resolution information; based on a matching result of the bill edge features and a preset template, geometric distortion correction is carried out on the original image, and a corrected bill image is generated; inputting the corrected image into a pre-training text region detection network to obtain positioning information containing text line boundary coordinates and region confidence; text line image blocks are extracted according to the positioning information, character segmentation preprocessing is executed, and a character-level image sequence is generated; calling a deep character recognition model to classify the sequence character by character, and generating an initial text recognition result; semantic verification and error correction are performed on the initial result based on the bill type knowledge base, final structured text data are generated, and the processing requirements of bills of different types and qualities can be met.

The invention discloses a long-distance binocular camera calibration optimization method based on multistage feature enhancement, and the method comprises the steps: carrying out the processing of a condition that a long-distance calibration plate has a highlight region and is fuzzy in angular points, employing a bilateral filter to suppress the reflection of a highlight mirror surface, reducing the reflection of light, and maintaining the definition of an image; the checkerboard texture is enhanced by adopting a CLAHE algorithm in combination with blocking processing and a contrast gain threshold value; a Sobel operator and a Harris corner response function are fused, gradient direction distribution characteristics of a pixel neighborhood are extracted through the Sobel operator, weighted fusion is carried out on the gradient direction distribution characteristics and the Harris corner response function, and the response intensity and specificity of a corner area are remarkably enhanced. A three-level image preprocessing framework including reflection suppression, contrast enhancement and corner enhancement is constructed, the problem of feature extraction of a traditional method under a complex illumination condition is solved, the corner area detection capability is enhanced, the problems of specular reflection noise, low contrast and corner blur are effectively solved, and the calibration precision is improved.

The invention discloses an infrared small target detection method and system, a detection device, an electronic device and a medium, and belongs to the field of image processing. A complete infrared image is input into a coarse and fine detection infrared small target detection framework, and the coarse and fine detection infrared small target detection framework screens a target area of the complete infrared image and then performs accurate target detection; obtaining an infrared small target detection result; the method comprises the following steps: inputting a complete infrared image into a region dichotomy network, carrying out dichotomy of an image block level, judging whether each image block contains a target or not, generating a multi-scale region feature map based on a judgment result, and in a lightweight target detection module, utilizing a convolutional neural network structure and combining a context-guided knowledge distillation module to detect the target. Carrying out different-level feature extraction and fusion on the multi-scale region feature map, and generating an infrared small target region detection result based on the combination of multi-level features; and carrying out region mapping on the infrared small target region detection result and the complete infrared image to obtain an infrared small target detection result.

The invention relates to the technical field of semiconductor packaging detection, in particular to a semiconductor packaging micro welding spot three-dimensional shape intelligent detection system which comprises a sample placing frame, an imaging module, an image processing unit, a detection module and a result display module. The image processing unit constructs a unified curvature field expression by fusing three-dimensional information of three different imaging principles of a spherical mirror model, a light-sectioning microscopic three-dimensional reconstruction algorithm and stripe structured light, and adopts adaptive mesh refinement to process a curvature change violent region; the detection module identifies welding spot defects such as pseudo soldering, bridging and poor connection based on differential invariants and singularity analysis, the system realizes high-precision detection of various semiconductor packaging forms such as BGA, QFN packaging and wafer-level packaging, the measurement precision is improved by 35%-40%, the processing efficiency is improved by about 50%, the defect omission ratio is reduced from 15% to below 3%, and the detection accuracy is greatly improved. And a brand new solution is provided for semiconductor packaging quality control.

The invention discloses a vehicle target motion state estimation method based on an unmanned aerial vehicle video, and relates to the technical field of computer vision and intelligent monitoring, and the method comprises the steps: receiving a real-time video stream of an unmanned aerial vehicle, decoding the real-time video stream to obtain an original video frame, carrying out the graying, zooming and normalization preprocessing to obtain a processed image, and synchronously adjusting the size of a vehicle bounding box; vehicle targets are detected, a unique ID is allocated for multi-target tracking, and a motion trail is maintained; determining a background feature region, detecting and tracking feature points to estimate an inter-frame transformation matrix, and performing accumulative transformation to obtain a background overall transformation relation and a historical frame-to-current frame mapping mechanism; calculating a displacement residual error based on the trajectory and the mapping, judging direction consistency, estimating a smooth speed, and updating a motion confidence coefficient to judge a vehicle motion or static state and static duration; outputting the ID, the state, the static duration and the average speed of the vehicle; the method effectively eliminates the motion interference of the unmanned aerial vehicle, improves the estimation precision, is suitable for the traffic monitoring of the unmanned aerial vehicle, and is high in real-time performance and reliability.

The invention relates to a remote sensing image change detection method based on frequency domain distribution alignment, a program, equipment and a storage medium, and the method comprises the steps: carrying out the preprocessing of two remote sensing images of different time phases in the same region, respectively carrying out the multi-scale feature extraction, obtaining a first feature map of each remote sensing image under each scale, and obtaining a second feature map of each remote sensing image under each scale; projecting to different frequency domains by using two-dimensional discrete wavelet transform to obtain sub-bands corresponding to the frequency domains; subtraction is carried out on the sub-bands of the two remote sensing images in the same frequency domain under each scale to obtain a frequency domain component of the scale; performing deconvolution on the frequency domain component according to a convolution kernel of two-dimensional discrete wavelet transform by adopting inverse wavelet transform, and fusing the frequency domain component into a second feature map of the scale; for each remote sensing image, splicing the first feature map and the second feature map under the minimum scale to obtain a fusion feature map of the remote sensing image; and respectively aligning the fusion feature maps of the two remote sensing images with the original input resolution, connecting the fusion feature maps and generating a change region detection map through a classifier, thereby realizing change region detection of the two remote sensing images of different time phases in the same region.

The invention discloses a skin image abnormal region detection method based on a convolutional neural network, and relates to the technical field of medical image analysis, and the method comprises the following steps: S1, carrying out adaptive illumination and color normalization processing and two-dimensional fast Fourier transform; s2, multi-scale representation is fused in a cross-scale mode; s3, learning the dynamic weight of the multi-scale features and carrying out weighted summation; s4, through improving a DANet model, executing double-path processing of Fourier domain semantic modulation and morphological prior space attention, and gating bidirectional aggregation; s5, carrying out binarization and connected domain analysis, and extracting candidate focus areas; s6, extracting an instance-level feature vector, and estimating a corresponding cognitive uncertainty value; and S7, performing graph relation reasoning and multi-head decoding. According to the method, the limitations of neglect of association between lesions, single evaluation dimension and poor prediction generalization ability in a traditional method are effectively overcome, and an efficient and reliable solution is provided.

Procedures, methods, architectures, apparatuses, systems, devices, and computer program products for near-field (NF) region detection. It may be beneficial if a wireless transmit-receive unit (WTRU) and the network are aware of whether the WTRU is within a NF region or not. A WTRU may be configured with a primary (P-RS) and a secondary (S-RS) set of resources. The WTRU performs measurements on each pair of associated RS resources selected from the P-RS and S-RS resource sets. Based on measurement results, the WTRU selects one or more RS groups that include a S-RS and an associated P-RS, and determines whether a pre-defined measurement condition is fulfilled. Based on the determined state for the selected RS groups, the WTRU determines to report an NF detection to the network, such as to enable location division multiple access and beam focusing which may enable enhanced spectrum efficiency.

The invention discloses a linear scanning tire image target detection method and system, and relates to the field of electrical digital data processing, and the method comprises the steps: a YOLOV8 network employs an FAE feature extraction module to replace original convolution, extracts channel features through depth separable convolution, and obtains a transverse scanning feature map through point-by-point convolution fusion. Inputting the feature map into a CCA mixed attention module, executing global maximum pooling to obtain a position code, and weighting with channel attention to obtain an adjustment weight feature map; and the shallow and deep features of the backbone network are spliced through a tiny target detection layer to form a fusion feature map, a detection head is input to execute rotation frame prediction, and a target frame containing a rotation angle is output. Angle classification loss and frame regression loss are calculated, a total loss value is obtained through weighted summation, and network parameters are updated through back propagation to complete training. And inputting the to-be-detected tire long image into the trained network to output a character region detection result. By implementing the method, the detection precision of the character target in the linear scanning tire image can be improved.

A system detects a traveling obstacle region that is a region that obstructs traveling of a work machine. The system includes a region angle detector, a region height detector, an angle obstacle region detector, a height obstacle region detector, and a traveling obstacle region detector. The region angle detector detects an angle of a traveling surface based on a distance image around the work machine. The region height detector detects height of the traveling surface based on the distance image. The angle obstacle region detector detects an angle obstacle region that is an obstacle region based on the detected angle. The height obstacle region detector detects a height obstacle region that is an obstacle region based on the detected height. The traveling obstacle region detector detects a traveling obstacle region based on the detected angle obstacle region and the detected height obstacle region.

The invention belongs to the technical field of medicine, and discloses a pulmonary nodule intraoperative positioning system based on a flexible array sensor, which comprises the flexible array sensor arranged at a to-be-detected part of a subject and used for collecting a compression signal and obtaining stress distribution; the nodule area detection unit is used for determining a nodule area through clustering analysis according to the stress distribution; the morphological optimization unit is used for performing morphological optimization on the determined nodule region; the feature extraction unit is used for extracting related geometric features based on the optimized nodule region; and the visualization unit is used for visually displaying the stress distribution, the optimized nodule region and related geometric features. The method can assist doctors in accurately positioning the pulmonary nodules, and meanwhile, quantitative features of the nodules are provided.

The invention relates to the technical field of image recognition, in particular to a multi-scale osteosarcoma CT image lesion area detection method based on feature distillation. The method comprises the following steps: acquiring a multi-scale osteosarcoma CT image and an attachment area image, performing forged image difference reduction to generate a unified CT image, reconstructing a three-dimensional tumor body form, extracting a bone attachment structure, performing relocation registration to obtain tumor body relocation data, and then generating a simulation attachment area framework based on the data, the method comprises the following steps: performing static feature distillation on a three-dimensional tumor body form to obtain an edge contour, defining an initial attachment area frame, performing feature comparison through a historical contour image to extract form change, evaluating the erosion degree of the attachment frame, and finally screening candidate lesion areas and positioning osteosarcoma lesion areas according to change areas. According to the invention, more efficient multi-source heterogeneous image data integration, more accurate three-dimensional tumor simulation reconstruction and clearer lesion area detection are realized.

The invention discloses an intelligent service system and method based on an intelligent home system, and relates to the technical field of intelligent home, and the system comprises an intelligent home equipment information management library, an intelligent home equipment analysis module, a scheduling analysis module and a home service analysis module. Equipment specifications, functions and operation parameters are acquired through an equipment information management library, and a function coverage rate, a cooperative gain and a scene adaptation degree are calculated, so that equipment capability comprehensive evaluation is realized; and the scheduling analysis module constructs a scheduling conflict and function dependency matrix, calculates a time adaptation rate, a function closed-loop rate and a conflict recovery rate in combination with connected region detection and closed-loop detection, forms a scheduling support degree, and realizes efficient resource scheduling. The home service analysis module constructs a life data model based on user operation and an equipment state, simulates multi-scene operation, and deduces an equipment satisfaction level; and finally, the display equipment is sorted according to the satisfaction degree, and adaptive and personalized matching conditions are marked for the user to select.

The invention provides a cowshed drivable area detection method based on fusion of a laser radar and a monocular camera, and the method comprises the steps: synchronously collecting real-time point cloud data and image data, carrying out the matching of the real-time point cloud data and a point cloud map constructed offline, and carrying out the calculation to obtain the pose state of a vehicle in a current map; the method comprises the following steps: inquiring and acquiring key points of a global induction area around a vehicle from a priori map constructed offline, and performing spatial mapping to form an image region of interest; pixel-level classification is carried out through a pre-constructed lightweight semantic segmentation model so as to output and obtain a binary segmentation mask; and mapping the binary segmentation mask from the image pixel coordinate system to the vehicle coordinate system through inverse perspective transformation, and generating a drivable area map under the vehicle coordinate system. According to the invention, through the core thought of priori map guidance, semantic fine recognition and coordinate system unified restoration, the drivable area detection of the cowshed is solved, and a basis is provided for a subsequent path planning module.

The invention belongs to the technical field of image processing, and particularly relates to a boundary intrusion target detection method based on machine vision, and the method comprises the steps: obtaining a boundary monitoring video frame sequence, extracting a grayscale image, and constructing a background model image based on a historical frame sequence; according to the gray scale change of the pixel point in the time window, a time sequence fluctuation index is obtained through the standard deviation of the time sequence gray scale set and the overturning frequency of the gray scale difference value; according to texture distribution characteristics in a pixel point neighborhood, obtaining a spatial discrete entropy through a gray level histogram; obtaining a direction chaos index through the vector sum modulus and the algebraic modulus sum of the motion gradient vector; and correcting the original motion response diagram according to the time sequence fluctuation index, the spatial discrete entropy and the direction chaos index to obtain an intrusion confidence coefficient, and carrying out connected region detection according to the intrusion confidence coefficient. According to the method, the environmental dynamic interference is effectively inhibited by fusing the time sequence oscillation, the space texture and the motion direction, and the accuracy of boundary intrusion detection is improved.

A warning system includes one or more processors configured to execute area detection processing of detecting an area of interest that is potentially unsafe for a user in a real space, execute distance determination processing of determining a distance between the user and the area of interest, and execute vibration generation processing of generating vibration that causes the user to sense a force in a direction away from the area of interest on a basis of a positional relationship between the user and the area of interest in a case where determination is made that the distance between the user and the area of interest is less than a threshold value.

The invention provides a multi-domain and Mama collaborative saliency target detection method oriented to a 360-degree image, mainly relates to a saliency region detection method oriented to image equatorial region structure modeling and global guidance enhancement, introduces PVT as a backbone network, extracts multi-scale features, inputs the multi-scale features to a frequency domain-space domain coordination module, and finally, obtains a multi-scale target detection result. The multi-scale features extracted by the PVT backbone are fully fused through frequency domain and spatial domain information, so that multi-scale edge details in the image can be effectively captured, and the significance boundary of the equatorial region is enhanced; an attention fusion Mama module is introduced, by fusing output features of a frequency domain-space domain coordination module, the Mama module can effectively improve structural guidance and semantic complementation of equator saliency information on a polar region, and finally a lightweight multi-stage feature aggregation module is designed for generating a saliency feature map. According to the detection method provided by the invention, the most advanced performance can be obtained under the condition of relatively low calculation complexity.

The invention discloses an unmanned aerial vehicle signal detection and identification method based on spectrum feature enhancement. The method comprises the following steps: firstly, carrying out time-frequency transformation on a received signal to obtain an original time-frequency graph; performing learning enhancement on the degenerated time-frequency graph by adopting a coding-decoding type deep neural network, and realizing noise suppression and structure recovery by combining pixel reconstruction, structural similarity and a texture perception loss function; and finally, performing target area detection and positioning on the enhanced time-frequency graph, directly outputting a structured result containing a time-frequency range, a category and confidence, and completing conversion from a frequency spectrum to a linkable engineering target. According to the unmanned aerial vehicle signal detection and recognition method based on spectrum feature enhancement, a spectrum feature enhancement mechanism is introduced before traditional spectrum analysis and feature recognition processing, and region-level detection and judgment are executed under the enhanced spectrum constraint condition; reliable discovery, positioning and identification of an unmanned aerial vehicle control link and an image transmission link in a complex electromagnetic environment are realized.

The invention discloses a sonar buoy layout simulation optimization method and electronic equipment. The sonar buoy layout simulation optimization method comprises the following steps: dividing a coverage area mark of a sonar buoy layout scheme into a plurality of grids; a population is initialized, the population comprises a plurality of chromosomes, each chromosome is initialized randomly, each chromosome comprises a plurality of groups of genomes, each group of genomes corresponds to one grid, each group of genomes comprises a plurality of genes, and each gene is used for representing a boundary adjustment value of the corresponding grid; a fitness function of the chromosome is set, and the fitness value of the fitness function is calculated based on the coverage rate, the overlapping rate and the balance degree of the number of sonar buoys of the grid; and carrying out iteration for multiple times until an iteration ending condition is met, ending the iteration, and outputting the optimal chromosome in the final population. The method effectively solves the problems of region overlapping and omission, and ensures the effectiveness of understanding and the global convergence capability of the algorithm. And the efficiency and the system response capability of multi-platform collaborative area detection in a complex sea area environment are remarkably improved.

The invention discloses a wide-spectrum rapid scanning and intelligent identification method, system and device, and the method comprises the steps: firstly, carrying out the rapid frequency sweeping through employing a low-sampling-rate and low-quantization-precision receiver, generating a low-resolution time-frequency graph, and recovering the super-resolution of the low-resolution time-frequency graph into a high-resolution spectrum through a pre-trained deep learning model; then, the enhanced time-frequency image is segmented into time-frequency blocks, global features are extracted through a Transform encoder, and detection and screening of multiple candidate signal areas are achieved; and finally, carrying out directional sampling on the candidate region, respectively extracting time domain and frequency domain features, mapping the time domain and frequency domain features to a unified space for fusion, and completing signal category judgment based on the combined features and a judgment network. According to the wide-frequency-spectrum rapid scanning and intelligent identification method, system and device, through a layered architecture of broadband rapid scanning-frequency spectrum quality recovery-multi-region detection-directional sampling and time-frequency fusion identification, the reliability of unmanned aerial vehicle signal detection and identification in a complex electromagnetic environment is improved while the hardware cost is reduced.

The invention provides a drivable area detection method and device and a vehicle, which can be applied to the field of intelligent driving. The method comprises the following steps: acquiring data acquired by a sensor; according to the data, information of a drivable area of the vehicle is determined, and the information of the drivable area comprises road attributes and / or boundary attributes of the drivable area. The method can be applied to an intelligent vehicle or an electric vehicle, the accuracy of planning and decision making of the vehicle can be improved, and the cost of intelligent driving can be reduced.

The invention provides an underground steel bar and cavity detection method and system based on a YOLO model, and the method comprises the steps: obtaining original image data for target region detection, carrying out the preprocessing of the original image data, obtaining the preprocessed image data, carrying out the marking of the preprocessed image data, and carrying out the detection of a target region. The method comprises the steps of obtaining an image data set marked with reinforcing steel bars and holes, taking YOLO as a basic detection model, carrying out multi-dimensional feature optimization on the YOLO model according to the characteristics of image data to obtain an optimized YOLO model, training the optimized YOLO model by using the image data set to obtain a trained target detection model, and detecting the target detection model according to the trained target detection model. And inputting image data of a to-be-detected target area into the trained target detection model, identifying a reinforcing steel bar and a cavity target, and outputting target information of the reinforcing steel bar and the cavity in the image. According to the detection method, efficient, high-precision and robust detection of underground steel bars and cavity targets is realized.