11results about How to "Improve perception accuracy" patented technology

The present application relates to the field of wisdom agriculture and image recognition technology, specifically to a crop disease and pest monitoring system based on image recognition, comprising: an environment perception module, which obtains the environmental context parameters of the monitoring area through multi-modal sensor nodes; a multi-source context analysis module, which constructs a disease and pest outbreak risk model in combination with historical disease and pest distribution spatio-temporal data and calculates the disease and pest risk level; an adaptive strategy generation module, which dynamically reconstructs the image acquisition mode and feature extraction strategy according to the risk level and the edge side resource constraint state, and generates adaptive perception instructions; an image acquisition execution module, which adjusts the physical acquisition parameters and algorithm calculation density of the visual sensor in response to the instructions, and executes image acquisition and recognition; the present application breaks the traditional rigid acquisition mode and realizes the optimal balance between monitoring efficiency and device survival period.

The invention provides vehicle glass, a vehicle glass assembly and a vehicle, the vehicle glass comprises laminated glass, a first bonding layer and a first glass plate, the first glass plate is connected with the laminated glass through the first bonding layer or arranged in the laminated glass through the first bonding layer, and the first bonding layer is arranged between the first glass plate and the laminated glass. The first glass plate comprises a first signal transmission area and a first shielding area, and the first shielding area is arranged around the first signal transmission area. Therefore, the problems of local optical deformation and distortion of the glass are reduced, and the diopter performance of the vehicle glass in a sensor area is remarkably improved.

This invention discloses a multi-dimensional privacy information perception and collection method. It constructs a formal description model of multi-dimensional privacy information, characterizing privacy information from multiple perspectives, including its composition, attributes, permissions, and control. Based on active learning and transfer learning algorithms, it builds a metadata intelligent perception, identification, and labeling method adaptable to various application scenarios and privacy protection strategies. This patent's advantages in adapting to multiple scenarios and information types allow it to adapt to the impact of changing application scenarios on privacy properties, reduce the concealment of privacy information caused by multiple information types, and improve the accuracy of privacy information perception. It focuses on perceiving multiple types of privacy information, comprehensively perceiving different privacy types such as text, images, and sound. Compared to existing single-type privacy perception methods, it is more beneficial for subsequent privacy strategy formulation and privacy control implementation, improving user privacy security. It has strong privacy perception capabilities, can process different types of data in parallel, and offers accurate privacy positioning and fast data processing.

This invention relates to a non-line-of-sight target detection method based on a smart metasurface on curved roads, comprising: setting up a smart metasurface device for relaying the detection of non-line-of-sight targets within a non-line-of-sight range; illuminating the smart metasurface device with a detection radar; and modulating the beam emitted by the detection radar based on the smart metasurface device to change the beam's propagation path and illuminate the non-line-of-sight target; the detection radar receiving the echo signal returned by the smart metasurface device and performing non-line-of-sight target detection and early warning based on the echo signal. This invention can accurately obtain the distance, speed, and angle information of non-line-of-sight (weak) targets, and has advantages such as all-weather and 24-hour operation, high sensing accuracy, wide sensing range, and low construction cost.

This invention provides a triaxial force sensor, comprising an elastic body and a strain gauge assembly. The elastic body includes connecting cylinders at both ends and a strain force detection area in the middle. The strain force detection area in the middle of the elastic body is configured as a weak connection structure, which includes at least four strain beams. The area enclosed by the strain beams is arranged in a square shape on the projection plane. The strain gauge assembly is attached to each strain beam. This invention can accurately measure forces in three mutually perpendicular directions and has the advantages of simple structure and fast response speed.

The application discloses a construction behavior anomaly detection method and system based on a lightweight large model, and relates to the technical field of artificial intelligence; the method comprises the following steps: collecting multi-source heterogeneous data of a construction site, including video streams, sensor data and building information model (BIM) data, and performing spatiotemporal alignment preprocessing; constructing a construction behavior analysis architecture based on a lightweight large model, utilizing a lightweight visual feature extraction network and a spatiotemporal graph convolution behavior recognition network to extract entity features, position information and action spatiotemporal features of personnel, equipment and materials; generating a fusion feature vector through a cross-modal feature fusion encoder combined with a self-attention mechanism; finally, utilizing a multi-modal large language model reasoning engine combined with a preset knowledge graph to perform logical reasoning, identifying the interaction relationship between personnel and the environment, thereby determining construction behavior anomalies and generating early warning information. The method effectively improves the accuracy and efficiency of construction site anomaly behavior detection, reduces the consumption of computing resources, and enhances the intelligent level of safety monitoring.

This invention discloses a soft measurement method based on PDA-GTCN-mBiGRU, belonging to the field of industrial process control and data processing technology. The method includes: constructing an input matrix containing the current time and historical time; constructing a PDA-GTCN-mBiGRU model, in which the upper branch first uses a gated temporal convolution module to extract temporal features, and then processes them through a parallel dual attention mechanism module; the lower branch uses a multi-time-step bidirectional gated loop module to extract global bidirectional dependency features; the features output from the upper and lower branches are concatenated and fused, and then weighted twice through a self-attention mechanism layer, finally outputting the predicted value through a fully connected layer. This invention, through the collaborative driving of the upper and lower branches, achieves a complementary fusion of local fine-grained feature capture and global temporal deep modeling, ultimately significantly improving the prediction accuracy and robustness of the soft measurement model when facing highly complex and strongly coupled industrial processes.

The application discloses a positioning method and device, a first node, a network function, a storage medium and a computer program product. The method comprises the following steps: a first node measures a first signal and a second signal to obtain a measurement result, the first signal comprises a reference signal transmitted through a first path, and the second signal comprises a reference signal transmitted through a second path; the first path comprises a transmission path directly from a second node to the first node or a transmission path from the second node to the first node after being reflected by a fixed object, and the fixed object is located at a preset position; the second path comprises a transmission path from the second node to the first node after being reflected by a target; the position information of the target is determined by using the measurement result, and the position information of the target is transmitted to a first function through a first communication interface; or the measurement result is transmitted to the first function through the first communication interface, so that the first function determines the position information of the target, and the first function is associated with a perception service.

This invention discloses a method for generating smooth trajectories for robotic arms that combines visual keypoint cues with a hierarchical diffusion model, applicable to complex robotic arm operations based on visual and verbal commands. The method constructs a hierarchical robotic arm operation framework. The high-level layer utilizes a sub-target generator guided by a visual-language model to predict the six-DOF pose sequence of the end effector as operation sub-targets from visual observations and verbal commands. The low-level layer proposes a dual diffusion model trained based on a random sampling strategy to generate smooth, low-curvature joint position trajectories between sub-targets. A keypoint extraction module that requires no training is introduced to provide visual cues. This invention improves the success rate of operations, reduces trajectory jitter and redundant motion, and enhances the smoothness and execution stability of robotic arm operations through the synergy of hierarchical planning and diffusion trajectory optimization.

The application discloses a kind of intelligent monitoring and control method and device for overflow type ball mill, belong to intelligent monitoring technical field of mine machinery.The device is through integration signal conditioning and adaptive acquisition hardware subsystem, edge inference calculation core of physical priori guidance and industrial control execution interface unit, real-time acquisition overflow type ball mill Multi-source perception signal, and through hardware closed loop control loop generation isometric vibration signal flow;Subsequently call fault mechanism mapping knowledge base, initialize the operation parameter of feature extraction circuit in each processing channel, obtain the decoupled component-specific feature vector;Finally, real-time calculation of physical state index, adaptive weighted fusion is carried out to component-specific feature vector, and the health state evaluation instruction and control instruction of key component are output in parallel in fault discriminator.The application can effectively overcome the problem that existing general monitoring equipment cannot adapt to speed drift, lack of edge intelligence and control response lag from hardware level.

The present application relates to the technical field of temperature control, in particular to a gradient temperature control method based on artificial intelligence, comprising: collecting multi-dimensional working condition parameters through a sensor array, constructing a dynamic characteristic matrix through cleaning and normalization; inputting the matrix into a long short-term memory network with attention mechanism, identifying time sequence correlation characteristics and time constant to construct a prediction model; mapping the characteristics to a virtual thermal dynamics simulation model for parallel simulation and rehearsal, evaluating energy distribution law and screening the optimal control increment sequence; converting the sequence into control signals to issue instructions to the actuator through the bus, executing active feedforward compensation; comparing the actual output with the simulation output in real time to calculate the residual error and dynamically correct the model parameters. The present application improves the forward-looking nature of temperature control decision-making through multi-dimensional feature extraction and virtual dimension rehearsal, improves the regulation stability of the controlled object under large-lag working conditions, and realizes precise active control of temperature gradient.