22results about How to "Target detection is accurate" patented technology

The invention discloses a 3D target detection method based on millimeter wave radar and camera image double-branch mutual fusion. According to the method, specific image features of respective modes are efficiently extracted from a camera and a millimeter-wave radar, and complementary branch information is introduced. In particular, the method generates geometrically perceived camera bird's eye view (BEV) features and semantically perceived millimeter wave radar BEV features. The two types of features are fused to form a comprehensively enhanced BEV feature map, and then through decoding processing, accurate 3D target detection is realized. According to the method, the accuracy and robustness of 3D target detection are remarkably improved, and experimental results show that the method can effectively detect and position the 3D target in the environment even under the conditions of complex scenes and severe weather.

The invention discloses a lightweight fabric surface defect detection method and device, and relates to the technical field of target detection. The method comprises the following steps of: performing data preprocessing and division on a pre-acquired fabric image data set; secondly, the model is improved, down-sampling Conv in a backbone and a check is replaced with ADown, StarBlock is introduced into a backbone part C2f to form a C2fStar module, and PCDeect is introduced into a head part of a detection head; the fabric defect picture data set is trained through the built YOLOv8 + Aown + C2fStart + PCDeect detection network, and a model pre-training weight is obtained; inputting a to-be-detected fabric picture into the optimal weight model for fabric defect detection, and outputting a detection result; the device comprises a processor, and when the processor runs the computer program, the steps of the lightweight fabric defect detection method are executed. According to the method, the parameter quantity and the calculation quantity are effectively reduced in a fabric defect detection task, meanwhile, the industrial defect detection requirement can be met, and the model precision is improved.

The invention provides an insulator defect detection method GAC-YOLO based on YOLOv7, and relates to the field of identification and detection of insulator targets. Comprising the steps that a GSConv module is optimized, and the calculation amount and the memory requirement are reduced while the precision is guaranteed; an ACmix attention mechanism is introduced, convolution and self-attention are fused, the prediction ability of the model is enhanced, and then the detection effect is improved; the centralized feature pyramid is fused, information exchange and integration can be achieved through feature maps of different scales, more comprehensive insulator features are further obtained, and therefore the insulator detection precision is improved. And performing an ablation experiment on each module of the model, and performing a contrast experiment with a mainstream deep learning model to obtain a final insulator detection model. According to the model realized by the invention, on the basis of ensuring the original detection rate of the YOLOv7, the accuracy of the model in insulator defect detection is improved.

The invention discloses a ship target detection method and system based on a large-kernel multi-scale attention mechanism. The ship target detection method comprises the steps of obtaining a to-be-detected ship target image; performing target detection on the to-be-detected ship target image by using a target detection neural network based on a multi-scale fusion strategy; in the target detection neural network, a detection task is divided into bounding box prediction, category prediction and object prediction, a large-kernel multi-scale attention module is integrated, the large-kernel multi-scale attention module firstly adopts 5 * 5 depth separable convolution to integrate local features, and then a two-dimensional large convolution kernel (k, k) is decomposed into cascaded horizontal depth separable convolution (1, 2); the method comprises the following steps of: extracting multi-scale information (k, k) and vertical depth separable convolution (k, 1), and weighting channel local features extracted by Ghost Module by utilizing the extracted multi-scale information; and the weighted features are deeply fused through the Ghost Module again, so that the accuracy of ship detection is further improved.

The invention discloses a ship target detection method based on an attention-enhanced feature pyramid network, and belongs to the technical field of remote sensing image processing. The method comprises the following steps: constructing an attention enhancement module; building an attention enhancement feature pyramid network based on an attention enhancement module and a corresponding ship target detection model; then ship target detection task training is carried out on the ship target detection model; and finally, performing ship target detection on a to-be-detected remote sensing image by using the trained ship target detection model. According to the method, the attention mechanism is fully utilized to capture global semantic information association more fully, fusion features have more complete multi-scale feature information, the model can extract data features more comprehensively on a remote sensing image ship target detection task, the image content is understood, and the target detection efficiency is improved. The method can achieve the accurate recognition and positioning of the ship target in a complex ocean background, and facilitates the improvement of the accuracy and robustness of remote sensing image ship target detection.

The invention provides a two-stage generalized small sample target detection method and system based on self-supervision, and relates to the technical field of deep learning, and the method comprises the steps: defining a potential knowledge container to store all potential knowledge, and defining a prototype container to store prototypes of various candidate boxes; dividing the training data set into a base class set and a new class set; the self-supervised category domain carries out pre-training and fine tuning on the detection model based on the potential knowledge of the base category set and the new category set; the self-supervised detection frame field carries out pre-training and fine tuning on the detection model based on the potential knowledge of the base class set and the new class set; and inputting a to-be-detected image to the detection model, obtaining features through the feature extractor, performing potential knowledge search and segmentation on the to-be-detected image based on the potential knowledge container to obtain a relationship between the to-be-detected image and each potential knowledge, and obtaining a candidate box category. According to the method, detection model training is carried out in the category domain and the detection frame domain by using self-supervised learning based on potential knowledge, and the method has good target detection capability for both a base class and a new class.

The invention relates to a target detection method and device for a hundred-million-level pixel fusion video, a hundred-million-level pixel calculation imaging system, computer equipment, a storage medium and a computer program product. The method comprises the following steps: acquiring multiple local videos shot by an array camera; performing size reduction processing on the local video image of the current frame, and inputting the local video image into a first target detection model for target detection to obtain a first detection result; performing moving target detection on the local video by adopting a frame difference method, and determining a target detection area meeting a small target size condition; inputting the image of the target detection area into a second target detection model for target detection to obtain a second detection result; and obtaining a target detection result of the current frame fusion video image of the fusion video based on the first detection result and the second detection result. By adopting the method, the target detection accuracy and detection efficiency of the hundred million-level pixel fusion video can be improved.

The invention discloses a sonar video target detection method and device based on a domestic sublimation AI processor, and belongs to the technical field of marine acoustic detection, and the method comprises the following steps: S1, building an edge intelligent hardware platform based on an Atlas200I A2 acceleration module; s2, designing a forward-looking sonar video target detection network, introducing a motion excitation module to enhance inter-frame motion features, and designing a self-attention feature aggregation module to aggregate time context features; s3, training the sonar video target detection network by using a two-stage training method, storing the weight and the structure of the trained network as an ONNX reasoning framework diagram file, and quantifying the weight process of the model by using an asymmetric quantization method; and S4, deploying a video target detection network by using a mercuric chloride CANN heterogeneous reasoning framework. According to the method, the challenge that a single-frame detector has poor detection on conditions such as motion blur and target hiding in a forward-looking sonar video is overcome, the detection precision is improved, and application and development of a localized AI processor in a forward-looking sonar detection technology are promoted.

The invention discloses a forest fire source detection method and system based on multi-source information fusion and a storage medium, and the method comprises the steps: carrying out the real-time data collection of a detection region, and enabling the real-time data to at least comprise infrared image data, visible light image data and temperature data; carrying out consistency and fusibility preprocessing on the real-time data; carrying out fire source key feature vector extraction on the preprocessed real-time data, wherein fire source key feature vectors comprise a heat radiation feature vector, a visual feature vector and a temperature abnormal point feature vector of a fire source; fusing the key feature vectors of the fire source, and generating a comprehensive feature vector; and fire source detection and positioning are carried out through the comprehensive feature vector. According to the invention, based on optical image data, infrared image data and temperature data, different data sources are fused by using a deep learning model; detecting and positioning a fire source through the comprehensive characteristics; therefore, efficient and accurate target detection is realized.

The invention relates to a power transmission channel hidden danger identification method and device, computer equipment and a storage medium, and relates to the technical field of electric power safety. The method comprises the following steps: acquiring a to-be-processed image of a power transmission channel; based on the trained target detection model, performing object detection on the to-be-processed image to obtain a first detection frame of a target tower contained in the to-be-processed image and a second detection frame of the target engineering machinery; performing object segmentation on the to-be-processed image based on the trained conductor segmentation model to obtain a plurality of target conductors corresponding to the target tower; projecting each target lead to the ground in the to-be-processed image to obtain a projection line segment, and determining a connection region of the projection line segment and the first detection frame as a region of interest; and determining a hidden danger identification result corresponding to the power transmission channel according to the position relationship between each second detection frame and the region of interest. The method can improve the recognition accuracy of the hidden danger of the power transmission channel.

The invention is suitable for the technical field of electric digital data processing, and provides a target detection method, device and equipment based on radar and video fusion and a medium, and the method comprises the steps: obtaining radar point cloud data and video data of a to-be-detected target; obtaining a target detection result based on a processing result of a radar and video fusion module in the improved YOLO model on the radar point cloud data and the video data; wherein the radar and video fusion module is used for preprocessing radar point cloud data and video data to obtain a radar image and a video image, and performing feature extraction, feature splicing, feature fusion, residual processing, secondary splicing and convolution processing on the radar image and the video image. According to the invention, radar and video data can be effectively fused, and the accuracy and robustness of target detection are improved.

This invention provides a target detection method, system, and vehicle based on multimodal pre-fusion, relating to the field of autonomous driving perception technology. The eye-sensing sensor module simultaneously acquires image data frames and laser point cloud data for the same area. Image data frames acquired at the same acquisition timestamp are correlated with laser point cloud data to generate correlated data frames. Uncorrelated image data frames and correlated data frames are then output to the domain controller. This correlation between image data frames and laser point cloud data at the sensor end reduces link time and ensures no loss of original information, guaranteeing the integrity of the sensor data. The domain controller performs target detection based on the uncorrelated image data frames and correlated data frames. The domain controller can perform target detection based on more accurate and timely correlated data frames, improving the accuracy of target detection.

The application provides an open vocabulary target detection method and system, which comprises the following steps: integrating a learnable prompt anchor module in a multi-scale feature layer of a preset cross-modal feature fusion network; injecting each prompt anchor point vector into a corresponding feature map through residual connection to dynamically correct the multi-scale semantic bias of the feature map; in the model training process, using a frozen CLIP visual encoder as a teacher model and the preset cross-modal feature fusion network as a student model to train a corresponding small target region mask, and calculating the distillation loss of the dynamically corrected feature map according to the small target region mask; outputting the multi-scale features corresponding to the feature map according to the distillation loss, and performing boundary box regression and open vocabulary classification on the multi-scale features through a preset detection head to complete the corresponding target detection. The application can effectively improve the detection efficiency of small region targets.

The invention discloses a data processing method for an AR intelligent camera device, and the method specifically comprises the steps: obtaining original image data shot by the AR intelligent camera device in real time, and carrying out the processing of the original image data through employing a self-adaptive illumination compensation algorithm, and obtaining first image data; performing denoising processing on the first image data to obtain second image data; whether any athlete in the second image data is shielded or not is judged, if yes, complementation is carried out according to the historical track of the athlete, and third image data is formed; marking illegal behaviors in the third image data to form fourth image data; and putting the fourth image data into an AR information display template for fusion processing to generate AR display image data. According to the invention, the ornamental value and interactivity of sports event live broadcast can be effectively improved, high-quality and personalized AR enhanced experience is provided for audiences, and the satisfaction degree and participation degree of users are greatly improved.

The invention discloses a YOLOv5-based hedge frame type grape bagging identification and positioning method, which is characterized in that grape target detection in a dynamic environment is carried out based on YOLOv5, a grape body surface three-dimensional model is constructed to acquire a grape space attitude and feed back three-dimensional coordinate value data of the grape space attitude, a bagging mechanism is guided to move, and the bagging mechanism can be used for bagging. Precise positioning of a target is achieved through fusion of depth information and image information, the feasibility of a bagging path and a bagging action is judged through a space volume calculation method, a bagging point and an opening binding point of the bagging action are estimated and analyzed, the success rate of the bagging action is increased, dynamic target detection is conducted through a YOLOv5 convolutional network, and the accuracy of the bagging action is improved. The grape target recognition accuracy and response speed are enhanced, the model is lightened through comparison of detection networks of the same type, dynamic detection is carried out faster and more accurately, and target operation is carried out in cooperation with a bagging mechanism.

The present application discloses a method and device for target detection in visible light and infrared images based on dual-stream attention, which relates to the technical field of target detection in visible light and infrared images. The visible light and infrared image data are aligned, and a 3×3 convolutional layer is used to preliminarily extract a shallow feature map. The shallow feature map is then input into a residual network composed of a dual-stream Transformer module to extract a deep feature map. The deep feature map, the shallow feature map and the original image are then aggregated, and a detection head is used to achieve target detection. The present application considers both channel and spatial attention, so that the target detection network model can not only simultaneously enhance the representation capabilities of target features in the channel dimension and the spatial dimension, improve the accuracy and robustness of target detection, but also better adapt to targets of different sizes and improve the generalization capability of target detection. Moreover, the present application comprehensively utilizes the advantages of visible light and infrared images to accurately extract features and achieve accurate detection in complex backgrounds, low contrast or occlusion conditions.

A target detection method, apparatus and device, and a readable storage medium relate to the technical field of target detection, and comprise the steps of obtaining high-level aerial view features corresponding to target laser radar data; performing two-dimensional convolution and pyramid pooling processing on the high-level aerial view features to obtain first aerial view features and second aerial view features; performing feature enhancement on the high-level aerial view features through two-dimensional convolution kernels of different sizes to obtain first enhanced aerial view features; performing feature enhancement on the high-level aerial view feature, the first aerial view feature and the second aerial view feature based on an attention mechanism to obtain a second enhanced aerial view feature; performing feature enhancement on the first enhanced aerial view feature and the second enhanced aerial view feature through two-dimensional convolution kernels with different sizes to obtain a target enhanced feature; and performing target detection based on the target enhancement feature to output a target detection result. The target detection efficiency can be effectively improved while the target detection precision is ensured.

The invention discloses a large-tilt image multi-scale form target detection method, which comprises the following steps: acquiring a to-be-detected large-tilt image, and preprocessing the to-be-detected large-tilt image; inputting the pre-processed large inclined image into the trained target detection model to obtain a final target detection result; wherein the target detection model comprises a first feature extraction structure, a shallow feature extraction structure, a deep feature extraction structure and an additional structure. According to the target detection method for the multi-scale form of the large-inclination image, the target detection model is adopted to perform target detection on the to-be-detected large-inclination image, and an additional structure is designed for the large-inclination image, so that the target detection of the large-inclination image is more accurate; through the staged training method, the difficulties of less data of the large inclined image and large target scale and form difference can be effectively solved.

This invention provides a SAR image aircraft target detection method based on scattering point sensing fusion, belonging to the field of remote sensing image processing technology. The method includes: acquiring a SAR image to be detected and preprocessing it to obtain a preprocessed SAR image; inputting the preprocessed SAR image into a pre-trained target detection network model to obtain aircraft target detection results; the pre-trained target detection network model includes a neck network, which is used to perform fusion processing on the input multi-scale feature maps using an asymmetric strategy; the multi-scale feature maps include: high-resolution feature maps, medium-resolution feature maps, and low-resolution feature maps; the asymmetric strategy is used to perform fusion processing at the fusion nodes of the high-resolution and medium-resolution feature maps using a scattering extremum preservation fusion module, and at the fusion nodes of the low-resolution feature maps using a convolution module. This achieves more accurate and interference-resistant aircraft target detection.

The embodiment of the application relates to a target detection network model training method based on pixel and point cloud feature fusion, comprising the following steps: acquiring a point cloud and an image of a driving environment of an engineering vehicle; projecting the point cloud onto the image to obtain a projection image; acquiring pixel channel features of the projection image; training an initial network model based on the pixel channel features to obtain a trained target detection network model. According to the method, the point cloud is bound with different numbers of pixel channels to obtain the projection image, and then the target detection network model is trained according to the projection image and the image features corresponding to the projection image, so that the target detection network model trained in this way can avoid errors caused by relative motion of sensors, and the target detection is more accurate.

The invention discloses a target detection method and apparatus, a computer device and a storage medium. The method comprises the steps of obtaining at least two frames of to-be-detected images; detecting each to-be-detected image by adopting a target detection model, and determining detection result information corresponding to each to-be-detected image; and performing fusion processing on the detection result information corresponding to the at least two frames of to-be-detected images, and determining a target detection result. According to the method, the detection result information of the at least two frames of to-be-detected images is fused, the detection result information of each frame of to-be-detected image can be comprehensively and accurately detected and analyzed, the target detection accuracy is improved, and the method has high application value.

The invention provides a target detection method and system, and the method comprises the steps: obtaining an original image corresponding to a target through a camera, and generating a corresponding BEV image according to the original image based on a preset rule; performing 2D target detection on the BEV image to generate a corresponding first detection result, and performing primary correction on the truncated target according to the first detection result; and performing 3D target detection on the BEV image based on the first detection result to generate a corresponding second detection result, and performing secondary correction on the truncated target according to the second detection result to perform target detection on the corrected truncated target. According to the invention, the central points of 2D detection and 3D detection can be corrected to the same position, and the working efficiency is correspondingly improved.