398results about How to "Suppress interference" patented technology

The invention relates to an Internet of Vehicles multi-agent cooperative sensing method and system. The method comprises the following steps: constructing a collaborative sensing network, wherein the collaborative sensing network comprises a self-agent and a plurality of collaborative agents; acquiring and processing sensing data through the collaborative sensing network; performing feature extraction to obtain intermediate features; self-adaptive sparsification is carried out to obtain sparse features, and the sparse features are compressed and transmitted to a self-agent; performing time sequence feature enhancement on the features of all the agents at the self-agent end; fusing the features to obtain fused features; and constructing an Internet of Vehicles perception model, and realizing perception by the detection model according to the fused features. According to the method, the calculation complexity of traditional global attention is reduced from the square level to the linear level through an adaptive sparsification mechanism, the calculation overhead is remarkably reduced while the multi-agent feature interaction precision is kept, and the method is more suitable for real-time operation on the vehicle-mounted edge equipment with limited resources.

The invention provides a PCB multi-type defect detection method and device based on frequency domain perception enhancement, a storage medium and a program product, and relates to the technical field of computer vision target detection.The method comprises the steps that a to-be-detected PCB image is input into a backbone network for shallow feature extraction, and shallow feature maps of different processing stages are obtained; respectively inputting the shallow feature maps into a multi-domain collaborative feature enhancement network, performing decomposition enhancement on the shallow feature map in each stage through wavelet transform to obtain an enhanced low-frequency component and an enhanced high-frequency component, and performing first reconstruction on the enhanced low-frequency component and the enhanced high-frequency component through first inverse wavelet transform to obtain an enhanced low-frequency component and an enhanced high-frequency component; obtaining a frequency domain enhanced feature map of each stage; inputting the frequency domain enhanced feature map into a neck network, and enabling the neck network to perform multi-scale feature fusion on the frequency domain enhanced feature map to obtain a fused feature map; and inputting the fusion feature map into a prediction output end for target prediction to obtain a PCB defect detection result.

The invention discloses a wafer defect detection method and system based on an attention guidance network, and belongs to the technical field of wafer defect detection, and the method comprises the following steps: 1, adjusting light supplement according to a detection demand, and then collecting a wafer image; 2, performing spectrum enhancement preprocessing on the acquired wafer image; and 3, extracting the features of the enhanced image through a multi-scale distributed feature extraction backbone network, and fusing the extracted features through an attention-guided feature pyramid network. Complex background interference of the wafer is effectively suppressed, and defect characteristics are remarkably enhanced; constructing a feature extraction mechanism capable of capturing local details and long-range context information at the same time; balanced and accurate detection of multi-scale defects is realized; the physical priori of the defect is embedded into the network in a learnable manner, so that the learning efficiency and generalization are improved; while ultrahigh precision is ensured, low model complexity is maintained, and the real-time requirement of a production line is met.

The invention discloses a deep sea cable fault accurate positioning method based on electromagnetic induction and transient characteristic analysis, and relates to deep sea electric power engineering. According to the technical scheme, the method comprises the steps of 1, signal excitation used for transmitting a signal source; 2, signal filtering is carried out to counteract geomagnetic background noise; 3, signal correction is carried out to solve the problems of modal aliasing and endpoint effect existing when a traditional EMD is used for processing complex signals; 4, signal feature extraction, wherein transient pulse features in the signals can be effectively enhanced; and step 5, fault positioning. The method is mainly used for non-intrusive fault detection of the power cable in the kilometer-level deep sea high-pressure environment, and high-precision fault point positioning is achieved.

The invention relates to the technical field of instrument and process measurement, in particular to a multi-range pressure gauge self-adaptive switching system and method based on time sequence feature perception, and the system comprises a data collection module which generates a smooth pressure time sequence signal in real time; the time sequence feature sensing module receives the smooth pressure time sequence signal and generates a pressure prediction trend signal of a future short-time window based on the pressure change trend feature; the self-adaptive decision-making module compares the pressure prediction trend signal with a preset current measuring range threshold value range; the switching execution module responds to the pre-switching trigger signal when the switching locking signal is in an unlocking state, and outputs the pre-switching trigger signal to the multi-range sensor group to complete switching of working channels; and the data fusion module performs weighted fusion processing on the two paths of output signals according to the switching process state signal provided by the switching execution module. According to the invention, the problem of switching lag of an existing pressure measurement system in the face of a pressure transient working condition can be solved.

The invention discloses a distributed photovoltaic panel anomaly detection method under a high-altitude view angle based on a YOLOv8n model, and belongs to the technical field of target detection. The model comprises the following steps: acquiring a distributed photovoltaic panel image data set under a high-altitude view angle, and performing data enhancement and labeling; an improved YOLOv8n model is constructed, and the improvement comprises the steps that a C2fAT module is introduced into a backbone network, and the small target feature extraction capacity is enhanced; an SPPF module is replaced by an SPPF-LSKA module, and complex background interference is suppressed; an EMA attention mechanism is introduced into the neck network, and the multi-scale adaptive capacity is improved; optimizing a training process by adopting a WIoU v3 loss function; and training and optimizing the model by using the training set, and finally outputting a detection result through the test set. According to the invention, the problems of false detection and missing detection of the distributed photovoltaic panel in a high-altitude view angle are effectively solved, the detection precision of the distributed photovoltaic panel is further improved, and inspection personnel are helped to troubleshoot the photovoltaic panel in an abnormal state in the high-altitude view angle.

The invention belongs to the field of multi-sensor fusion systems, and discloses a multi-sensor fusion system simultaneous positioning and mapping method and a related device, and the method comprises the steps: obtaining an environment image, a laser point cloud and IMU data; scene semantic information is extracted through a multi-modal large model; dynamically generating a weight matrix of vision and laser radar relative to an IMU (Inertial Measurement Unit) by utilizing a vision and point cloud condition deep neural network in combination with scene information and IMU observation; and performing local and global optimization by combining each sensor speedometer factor, an IMU pre-integration factor and the weight matrix, and finally outputting robot state estimation and map point coordinates. Through semantic understanding and a self-adaptive weight mechanism driven by a conditional deep neural network model, a fusion strategy can be intelligently adjusted before environment change or sensor degradation, the robustness, precision and consistency of a system in an unstructured scene are improved, and the problems of positioning drift and map failure caused by fixed weight in a traditional method are solved.

The invention discloses a financial credit data verification system with zero knowledge proof, and relates to the technical field of financial science and technology. According to the system, a ZKP multi-source credit data acquisition module acquires multi-dimensional credit data of a borrower, and a data available and invisible mechanism is adopted to protect privacy; the credit data entropy analysis module quantifies data uncertainty based on a clustering algorithm, dynamically generates a threshold value in combination with a historical entropy value, and quickly screens abnormal data; a ZKP enhanced feature extraction module captures a dynamic change rate index and zero knowledge proof features; the deep verification model evaluation module utilizes a deep belief network and a federated learning framework to realize high-precision anomaly recognition on the premise of protecting privacy; the verification strategy regulation and control module generates a parameter regulation and control strategy through a swarm intelligence optimization algorithm; the feedback optimization module dynamically adjusts parameters through a closed-loop feedback mechanism, and improves the efficiency, precision and safety of financial credit data verification.

The invention relates to the technical field of electromagnetic shielding cabinet testing, in particular to a dynamic performance testing device and method for an electromagnetic shielding cabinet, and the device comprises an environment testing cabin which is used for placing a to-be-tested cabinet and providing a controlled electromagnetic background environment; the dynamic load simulation system is mechanically connected with the to-be-tested cabinet and is used for applying physical excitation with preset strength to the to-be-tested cabinet so as to simulate structural deformation of the to-be-tested cabinet in a working state; and the signal transmitting system comprises a transmitting antenna arranged outside the to-be-tested cabinet and is used for generating an electromagnetic signal covering a preset frequency band. The dynamic load simulation system simulates vibration, cabinet door opening and closing and other actual dynamic working conditions, the temperature and humidity adjusting module is combined to simulate different climatic environments, meanwhile, the high-speed synchronous triggering module is used for achieving high-precision synchronization of dynamic load applying, electromagnetic signal transmitting and data collecting, multi-working-condition collaborative simulation and precise synchronous testing are achieved, and the testing efficiency is improved. And the authenticity and the reference value of a test result are greatly improved.

The present application relates to the field of intelligent traffic parking scheduling, and particularly relates to a parking resource comprehensive scheduling strategy optimization method and system based on multi-region cooperation, which specifically as follows: collecting multi-region and multi-type parking resources, commuting parking demand, illegal parking risk and road network operation data according to scheduling period, constructing parking cooperative monitoring sample through standardization processing; calculating time-sharing sharing potential index, underground priority access intensity and road overflow risk elasticity index, constructing hierarchical directed ant colony search graph with heuristic information written; introducing capacity constraint self-repairing mechanism and congestion degree pheromone updating mechanism, solving Pareto solution set through multi-objective ant colony optimization, screening optimal compromise scheme combining fuzzy membership and dynamic preference weight, outputting parking guidance instruction and realizing rolling closed-loop iteration. The present application can realize time-sharing cooperative deployment of parking resources in residential and office areas, balance various parking load, and improve overall utilization efficiency of parking resources.

This invention relates to the field of atmospheric particulate matter analysis and discloses a deep learning-based method for calibrating organic aerosol concentrations. This invention addresses the technical problems in existing organic aerosol concentration calibration techniques, such as insufficient utilization of multi-source information, inadequate use of temporal evolution information, and fixed feature fusion methods lacking dynamic weight adjustment. By leveraging multi-source observation vectors, it fully utilizes observation information from different physical properties to establish a stable nonlinear mapping relationship under complex atmospheric conditions and mixed aerosol backgrounds. Through multi-source observation features and historical state features, the model can characterize the dynamic evolution of organic aerosol concentrations. Employing a bidirectional long short-term memory network and SE attention mechanism to process multi-source observation features and historical state features, it can automatically learn and dynamically adjust the importance weights of feature channels, effectively suppressing noise and redundant feature interference. Furthermore, it utilizes forward and backward time-series modeling to fully capture the contextual relationships between consecutive time points.

This application relates to the field of pipeline network detection technology, and discloses a visual detection method and system for pipeline network node defects based on a detection robot. The method includes: acquiring node images of a sewer network using a detection robot, and performing circle detection and sector segmentation on the node images to obtain multiple sector region blocks; extracting the block feature vector of each sector region block; calculating angular attenuation weights and radial attenuation weights, and constructing a node connection relationship matrix based on the angular attenuation weights and radial attenuation weights; combining the block feature vectors and the node connection relationship matrix to perform feature aggregation and illumination compensation to obtain a global feature vector; calculating angular deviation index, radial mutation rate index, and leakage index, and generating defect visual detection results, thereby realizing comprehensive diagnosis and spatial positioning of pipeline network node connection defects, and improving the defect detection accuracy of sewer detection robots in complex node environments.

The invention discloses a voiceprint detection-based upender fault detection method and system. The method comprises the following steps of: acquiring a sound signal in an operation process of an upender and preprocessing the sound signal; performing framing and fast Fourier transform to obtain frequency domain spectrum data; obtaining a Mel spectrum through a Mel filter bank, performing logarithmic operation on the Mel spectrum to obtain a logarithmic Mel spectrum, and extracting voiceprint features; performing normalization processing and splicing to obtain a voiceprint feature sequence to be detected; establishing a normal voiceprint model and a fault voiceprint model to form a voiceprint model library; using the improved RealNVP model to output a matching result with the voiceprint model library, and obtaining a running state category result; according to the invention, the operation state monitoring and fault identification of the upender are realized, and the stability and accuracy of fault detection are improved.

The invention belongs to the technical field of trace detection and intelligent early warning, and particularly discloses an intelligent sensing fire early warning system based on pyrolysis particulate matter detection. The system comprises a multi-angle polarized light scattering spectrum acquisition device, a physical information neural network inference engine, an airflow interference filtering unit, a pyrolysis dynamics constraint embedding module and a fire trend early warning output unit. A pyrolysis dynamic rule in an Arrhenius equation form is used as a physical constraint to be embedded into a neural network loss function, air flow disturbance is filtered out by combining topological data analysis and computational fluid mechanics, and a pyrolysis reaction stage and a fire behavior increasing trend are reversely deduced by using a multi-angle polarization scattering spectrum. And graded early warning is realized before the particulate matter concentration does not reach a traditional threshold value. According to the technical scheme, the accuracy, the robustness and the response lead of early fire recognition are improved.

This invention discloses a multi-source electromagnetic noise suppression method based on noise classification and deep learning, belonging to the field of geophysical electromagnetic exploration technology. The method includes: segmenting and preprocessing the original electromagnetic observation sequence; using an improved U-Net network with an encoder embedded in a Mamba time-series modeling module for low-frequency noise suppression; identifying strong noise types in the data segments using a ROCKET classifier; based on the classification results, calling a second improved U-Net network trained for the corresponding noise type for class-based targeted denoising; and finally, splicing the data segments to obtain complete, high-quality data. This invention, through a phased processing framework of "low-frequency pre-suppression—noise classification—class-based denoising," combined with the strong time-series modeling capabilities of the Mamba module and the efficient classification performance of ROCKET, significantly improves the suppression accuracy and signal fidelity for complex, multi-source electromagnetic noise, and is particularly suitable for processing ground and airborne electromagnetic exploration data.

The invention provides a steel plate defect detection method and system based on machine vision, and relates to the technical field of steel plate defect detection.The method comprises the steps that upper and lower surface sensors, an electrical cabinet, a detection server and a camera computer are deployed on a production line, and power supply connection is conducted through the electrical cabinet; irradiating the steel plate by using an LED light source, and triggering a line scanning camera by an encoder to acquire upper and lower surface and edge images; processing and filtering pseudo defects through a camera computer, identifying real defects and extracting images, and determining defect types, positions and severity by combining a multi-level classifier and an expert database; after related data are stored, quality files such as coiled material reports are automatically generated, sound-light alarm is triggered when the defect severity threshold value is exceeded, the field control unit is linked to execute feedback operation or output I / O signals, online detection, classification, storage, report generation and threshold-exceeding alarm linkage of steel plate defects can be achieved, and the detection efficiency and the quality control level are improved.

The application relates to the technical field of motor assembly quality image detection, and discloses a visual-based motor wire harness connection part abnormality detection method and system, wherein the method comprises the following steps: acquiring a motor wire harness connection part image; constructing a connection part illumination state vector; performing adaptive brightness correction; suppressing reflection and pollution interference; determining an effective wire harness connection part area; and outputting a wire harness connection abnormality detection result. Compared with the method in the prior art which mainly relies on single detection confidence for judgment, especially under complex working conditions such as strong reflection of a metal shell, oil stain adhesion and wire harness shadow superposition, the technical problem that it is difficult to stably distinguish between real structure abnormality and illumination artifact interference is solved. Due to the introduction of adaptive brightness correction and the joint mechanism based on attention enhancement target detection and assembly geometric constraint, the accuracy of motor wire harness connection part abnormality detection is improved.

The invention discloses a stereoscopic warehouse material inclination detection method based on a two-dimensional image and related equipment. An expensive three-dimensional sensor or a depth camera is replaced by a common camera, so that the equipment cost is greatly reduced, and popularization and application in a large-scale warehousing system are facilitated; angle correction is carried out through the preset installation angle, the problem that the relative angle between the camera and the goods shelf is not fixed when the camera is installed on the mobile storage equipment is effectively solved, and the robustness and adaptability of inclination detection are improved; an edge detection and robust IQR abnormal value elimination and length weighted average statistical method is adopted, so that noise and abnormal value interference are effectively suppressed, and the detection precision is greatly improved; adaptive detection of various goods allocation states is supported, detection modes are automatically switched according to different goods allocation states, invalid detection is avoided, and the overall operation efficiency and the resource utilization rate of the system are improved; and a confidence evaluation mechanism is introduced to comprehensively evaluate the credibility of the detection result, so that a reference basis is provided for subsequent decision making, and the false alarm rate is reduced.

The utility model provides an oil-resistant high-precision embedded thermal resistor, which belongs to the technical field of embedded thermal resistors, and comprises a reference end and a measuring end, the measuring end comprises a pipe sleeve head and two resistor elements arranged in the pipe sleeve head, the upper part of the pipe sleeve head is welded with an armored lead wire through laser, and the armored lead wire is connected with the reference end and the measuring end. The armored lead is sleeved with a silica gel heat shrink tube or a polytetrafluoroethylene tube, and a shielding lead is arranged between the armored lead and the reference end. According to the embedded thermal resistor, double resistor elements are adopted, so that the redundancy of a product is increased, cross validation can be realized, the measurement error is reduced, meanwhile, the lead directly connected with the measurement end is an armored lead, and a silica gel heat shrink tube or a polytetrafluoroethylene tube is sleeved on the lead, so that the oil resistance of the embedded thermal resistor is greatly improved, and the service life of the embedded thermal resistor is prolonged. In addition, environmental interference can be inhibited, thermal stability can be maintained, the service life can be prolonged, and the effective precision under actual working conditions is remarkably improved, especially in high-electromagnetic-interference, severe-temperature-change or corrosive environments.

The application discloses a YOLO-based avalanche monitoring self-improvement system and method, and belongs to the technical field of intelligent geological disaster monitoring. The method collects multi-source image data through a UAV and satellite remote sensing, and realizes initial avalanche detection through YOLO. A time sequence consistency verification mechanism is used to screen high-confidence pseudo labels, and a dynamic time attenuation weight strategy is combined to construct an enhanced training set. When the pseudo labels accumulate or the performance decreases, model retraining is automatically triggered, performance evaluation is used to decide model updating or rollback, and a self-learning closed loop is formed. The application innovatively deeply integrates YOLO detection, automatic pseudo label distillation, physical parameter inversion and edge end spread deduction, realizes the whole-chain self-improvement of detection-deduction-warning-optimization, continuously improves the detection accuracy, reduces the artificial labeling amount, supports complex terrain and extreme environments such as strong reflection and dynamic blur, and reduces the early warning delay to the millisecond level.

A laser-assisted positioning mechanism for spinning belongs to the technical field of mechanical processing. The mechanism aims to solve the problems of low precision and long calibration time in the traditional positioning method of multi-axis numerical control spinning equipment. It includes a sliding table module composed of a sliding table support and a sliding table body. The sliding table body is connected to the sliding table support through a linear guide rail. A servo motor drives the displacement of the sliding table body through a synchronous belt transmission mechanism, which includes a first synchronous wheel, a second synchronous wheel, and a closed-loop synchronous belt. The bottom of the sliding table body is fixed to the synchronous belt. A laser level is rigidly fixed to the side of the sliding table body through four symmetrical mounting holes and a mounting slot, projecting a laser reference line parallel to the axis of the spinning wheel. The bottom of the sliding table support is provided with a double fixed seat, which is rigidly connected to the processing platform through bolts. The mechanism realizes a displacement accuracy of 0.1 millimeters through closed-loop control of the servo motor, allowing the laser reference line to be dynamically adjusted along the axis of the spinning wheel.

The embodiment of the application relates to the technical field of communication, and discloses an antenna array and a communication device, the antenna array comprises a bearing base body, a waveguide antenna unit, a choke component and a feeding structure, the waveguide antenna unit comprises a plurality of first antenna elements and a plurality of second antenna elements, the first antenna element comprises a first rectangular waveguide cavity and a first radiation aperture arranged on the first rectangular waveguide cavity, the second antenna element comprises a second rectangular waveguide cavity and a second radiation aperture arranged on the second rectangular waveguide cavity, and the choke component is arranged on the waveguide antenna unit. In the above manner, the embodiment of the application can effectively constrain electromagnetic waves to propagate in a specific path, can significantly improve energy transmission efficiency compared with an open radiation structure, effectively suppresses array interference, and provides a high-performance antenna for a 72-78 GHz frequency band communication system.

The invention relates to the field of ship energy control, in particular to an electric ship energy control method based on working condition recognition. The method comprises the following steps: acquiring planned route information, and analyzing historical working conditions of route segments one by one to obtain historical working condition characteristics; real-time working condition parameters in the ship navigation process are collected; positioning and matching a plurality of sub-legs according to the real-time working condition parameters, and constructing a sub-leg working condition feature set; working condition deviation comparison is carried out on the historical working condition features according to the sub-leg working condition feature set, and a deviation state sequence is determined; performing propulsion power gap calculation based on the deviation state sequence to obtain a power gap compensation value; battery capacity information is identified, current endurance state evaluation is carried out, and an endurance evaluation value is obtained; and carrying out endurance constraint correction on the power gap compensation value according to the endurance evaluation value to obtain a power constraint adjustment value. According to the invention, through real-time working condition identification and analysis, energy efficiency optimization, endurance safety and operation reliability are considered under complex navigation working conditions.

The application belongs to the field of medical image recognition and relates to an esophageal lesion image recognition method based on a deep neural network, which comprises the following steps: step 1, processing actual digestive endoscopy video data to construct training data; step 2, constructing a deep neural network model and training the deep neural network model through the training data to obtain an early esophageal cancer recognition model; step 3, inputting digestive endoscopy video data to be detected into the early esophageal cancer recognition model, and the early esophageal cancer recognition model judging whether there is a suspected esophageal cancer lesion and outputting a lesion judgment result; the lesion judgment result comprising a spatial position and a progression stage of the lesion; the method models the time sequence dynamics of lesion characteristics in the endoscopy recognition process, and combines a graph neural network to depict the feature correlation of early esophageal squamous cell carcinoma under multi-view conditions, so as to improve the recognition accuracy, consistency and robustness of the model in a real clinical application scenario.

The application discloses a cross-cultural communication non-verbal media emotion recognition interaction system, which comprises a non-verbal media data acquisition module, a non-verbal emotion feature extraction module, a non-verbal emotion recognition model construction module, a cross-cultural feature weight parameter self-adaptive acquisition module and an emotion intelligent interaction module. The application relates to the technical field of data processing, in particular to a cross-cultural communication non-verbal media emotion recognition interaction system, which proposes a hierarchical collaborative design of a general non-verbal emotion recognition model and a cross-cultural feature weight parameter self-adaptive acquisition module, can significantly improve the accuracy of non-verbal emotion recognition in a cross-cultural scene, improves the recognition accuracy of the model by designing a three-dimensional structured model input tensor, a customized emotion feature reinforcement filter and an emotion fluctuation extraction pooling operation, and improves the accuracy of cross-cultural non-verbal emotion recognition by adopting a stage-by-stage position updating strategy in a predation stage and combining a dynamic dimension exchange mechanism to improve and optimize the algorithm.

This invention relates to the field of image generation technology, and more particularly to a method for generating bird's-eye view images of reservoir landscapes based on a Stable Diffusion large model. The method includes the following steps: preprocessing acquired reservoir landscape images and semantically annotating them based on water conservancy terminology to construct an image-text pairing dataset; training a pre-trained Stable Diffusion model, incorporating LoRA adaptation parameters into its UNet attention layer, using the image-text pairing dataset to generate a target feature image; and updating the LoRA adaptation parameters based on the difference between the target feature image and the reservoir landscape image to obtain a LoRA model for the reservoir landscape bird's-eye view image. This invention achieves structurally controllable and semantically consistent reservoir landscape generation by fusing structural reference constraints and semantically weighted alignment and introducing LoRA modulation.

The application relates to the technical field of LED circuit, in particular to a lamp color adjusting circuit with a thyristor light modulation module, which comprises a rectifier circuit, a multi-grade dial code color adjusting circuit, a frequency flash removing circuit, a thyristor current discharge circuit and a constant current light modulation circuit. The rectifier circuit converts an alternating current power supply into a direct current voltage to supply power to a circuit system. The multi-grade dial code color adjusting circuit adjusts the current distribution of LED lamp beads through a dial code switch and a second group of resistors, generates a color adjusting signal, carries out low-pass filtering and dynamic compensation on the color adjusting signal, and outputs a pure control signal to the thyristor current discharge circuit. The thyristor current discharge circuit adjusts a current path according to the control signal to provide a smooth and adaptive power supply for the constant current light modulation circuit. The constant current light modulation circuit realizes constant current output. The design significantly reduces the hardware cost and the complexity of the circuit system, effectively suppresses electromagnetic interference in combination with the low-frequency modulation characteristics of the thyristor, and realizes the balance between light modulation precision and the stability of the circuit system.

This invention belongs to the field of computer vision and image processing technology, and particularly relates to a query key-guided asymmetric feature enhancement system and method. The system includes: Q-ASFE modules embedded in the backbone network, specifically in the link where features are transmitted from the C3K2 module to the CBS module. The P2 and P3 layer feature maps output by the backbone network are processed by the neck network to output three feature fusion maps at different scales. These three feature fusion maps are then processed by a detection head to obtain the final detection result. This invention achieves efficient enhancement of image features through the organic integration of spatial detail awareness and global semantic guidance.

The invention discloses a laboratory environment regulation and control auxiliary system based on artificial intelligence. The system comprises a data acquisition module, a laboratory scene recognition module, a laboratory environment suitability evaluation module, a laboratory environment regulation and control strategy generation module and an environment intelligent regulation and control auxiliary module. The invention relates to the technical field of data processing, in particular to a laboratory environment regulation and control auxiliary system based on artificial intelligence, which innovatively introduces a laboratory scene recognition module to recognize a laboratory scene in real time and effectively improve the environment regulation and control precision; a clustering center based on mahalanobis distance is introduced to initialize, a laboratory scene recognition objective function fused with prior reinforcement constraints is constructed, a weak correlation clustering center recognition strategy and membership redistribution optimization are used for carrying out clustering algorithm improvement, and the accuracy of laboratory scene recognition is improved; and a historical optimal position memory mechanism is introduced, and a dimension learning neighborhood search strategy is adopted to improve an optimization algorithm, so that the precision of a laboratory environment regulation and control strategy is improved.

The application discloses a prior feature auxiliary directional coordinate attention remote sensing road extraction method, and relates to the technical field of remote sensing image processing.The method comprises the following steps: extracting multi-source prior features representing road physical characteristics and optimizing and fusing weight to obtain a fused prior feature map; processing the RGB remote sensing image and the fused prior feature cooperatively to form a fused feature containing multi-scale semantic information and physical prior; predicting a pixel main direction angle in the feature through a directional coordinate attention mechanism, pooling the feature along the main direction and the orthogonal direction to generate an attention map and re-calibrate the original feature; performing step-by-step up-sampling on the decoder and introducing a deep supervision mechanism; and training the model by using a composite loss function containing a Dice loss, a Focal loss and a boundary perception loss.The application fully utilizes road physical characteristic prior knowledge, combines the directional coordinate attention mechanism to capture road geometric characteristics, and effectively improves the accuracy and integrity of remote sensing road extraction in a complex scene.