2088results about How to "Realize detection" patented technology

The invention discloses a traffic incident detection system based on multi-modal data, and belongs to the technical field of traffic incident detection. The invention aims to solve the problems of poor traffic event detection capability and incapability of effectively associating context information caused by insufficient video data in a training process of an existing traffic event detection method. Comprising the following steps: acquiring high-dimensional visual features of a sample event image and a sample event video by adopting a visual encoder; obtaining a text feature vector of the cue word description text of the sample event by adopting a text encoder; a feature mapping layer is adopted to obtain high-dimensional visual features aligned with the text feature vectors; common multi-modal features are obtained through a multi-modal joint learning module; obtaining a semantic prediction result and an answer type text sequence of the sample event through a large language module; and obtaining a traffic event classification result through a traffic event classification module. The method is used for on-line detection of traffic events.

According to the harmonic amplification detection method and system of the electric power system, the frequency response of the impedance of the electric power system containing inter-harmonic frequency components is analyzed by obtaining the topological structure and the load distribution condition of the electric power system, and the impedance frequency characteristic scanning is performed on the electric power system at different frequencies by adopting a harmonic injection method; obtaining an impedance frequency response curve; according to the frequency interval of harmonic amplification, the hazard coefficient of the harmonic is evaluated through the voltage distortion rate and the current distortion rate caused by the harmonic, and the type of harmonic source equipment causing harmonic amplification is determined according to the hazard coefficient. According to the scheme, the harmonic source equipment can be detected, the power system can be conveniently maintained in time, and the stability of the power system can be improved.

The invention relates to the field of welding seam detection, in particular to an air conditioner refrigeration copper pipe welding defect detection device which comprises a base, positioning assemblies are symmetrically arranged on the two sides of the upper end of the base, each positioning assembly comprises a shell with a central through hole, a to-be-detected copper pipe horizontally penetrates through the central through hole of the shell, and a fixed cross beam is fixedly arranged between the top ends of the shells. Fixed ring frames are fixedly arranged at the two ends of the bottom of the fixed cross beam, movable ring frames are rotatably mounted in the fixed ring frames, the two movable ring frames are fixedly connected through mounting plates, the number of the mounting plates is four, the mounting plates are distributed in a cross shape, a pressure assembly is arranged on the mounting plate at the top end, and a detection assembly is arranged on the mounting plate at the bottom end. The stress load is gradually applied to the weld joint position of the to-be-detected copper pipe through the pressure assembly, and the deformation process of the weld joint position is monitored in real time in cooperation with the detection assembly, so that the bending strength of the weld joint position of the to-be-detected copper pipe can be dynamically detected in the circumferential direction.

The invention provides a welding defect identification method, system and equipment for an automobile part and a medium, and relates to the technical field of automobile welding.The method comprises the steps that a structured light vision sensor and a high-frequency ultrasonic sensor are used for obtaining three-dimensional structured light data of a to-be-welded area of the automobile part and sound wave signals generated in the welding process; obtaining characteristic frequency information used for identifying welding internal defects; the method comprises the following steps: capturing a heat distribution image of a to-be-welded area of an automobile part through a thermal imaging camera to obtain area feature information used for identifying welding defect thermal anomaly, and obtaining magneto-optical signal feature information used for identifying welding defects; performing defect identification and classification on the fused feature representation by training a deep learning model to obtain a defect identification result; and welding parameters are adjusted through a simulated annealing algorithm, so that the welding defects of the automobile parts are recognized. According to the invention, various welding defects can be identified, so that the accuracy and efficiency of welding quality detection are remarkably improved.

The embodiment of the invention provides a synchronous code detection method, system and device in a low earth orbit satellite communication system, and the method comprises the steps: grading an obtained maximum frequency offset range value of the low earth orbit satellite communication system, and obtaining frequency offset values of different grades, performing synchronous code capture on a first receiving signal of the input signal flow under the frequency offset values of different grades, and obtaining an initial frequency offset estimation value according to a capture result; compensating a second receiving signal of the input signal flow by using the initial frequency offset estimation value, and performing synchronous code capture on the compensated second receiving signal to obtain a candidate synchronous code position confirmed by coarse compensation; based on the candidate synchronous code position, performing frequency offset estimation on the synchronous code data under the plurality of time slots to obtain a candidate frequency offset estimation value; and compensating a third receiving signal of the input signal flow by using the candidate frequency offset estimation value, and performing synchronous code capture on the compensated third receiving signal to obtain a synchronously captured target synchronous code position. According to the invention, the detection of the synchronous code in the low-orbit satellite communication system is realized.

The invention relates to the technical field of bridge pier damage detection, in particular to a multi-source data fusion bridge pier collision damage detection device and method, and the device comprises piezoelectric sensors which are distributed in a bridge pier in an array; the surface crack detection device comprises a vertical walking bracket and an annular walking mechanism; the annular walking mechanism comprises an annular track, a rotator is arranged in the annular track, and an annular driving system is further arranged on the annular track; an image acquisition device which rotates along with the circumference of the rotator is also arranged on the rotator; the device further comprises a signal processing system which is in signal connection with the piezoelectric sensor, the vertical walking support, the annular driving system and the image acquisition device. According to the method, the surface crack of the pier can be detected, the internal damage of the pier can be detected, and various detection data are fused, so that the damage state of the pier after being impacted is comprehensively determined, and the accuracy and reliability of pier damage evaluation are improved.

The utility model discloses a tensile test device for detecting the toughness and the tensity of a film, and belongs to the field of film detection devices. A tensile test device for detecting toughness and tensity of a film comprises a base, an I-shaped frame, a mounting frame and scale marks are sequentially fixed to the upper side of the base, and a bidirectional driving mechanism is fixed to one end of the mounting frame; according to the utility model, through the cooperation of the bidirectional driving mechanism, the clamping assembly, the reciprocating driving assembly, the pressing roller and the scale marks, the comprehensive detection of the toughness of different parts of the film is realized, the bidirectional driving mechanism can control the opening degree of the film, and the reciprocating driving assembly enables the pressing roller to carry out pressing detection at different positions of the upper surface of the film; in addition, due to the arrangement of the scale marks, a detection part can be accurately positioned conveniently, the tension and pressure sensor is combined to monitor the stress condition of the thin film, the toughness and the tension degree of the thin film can be detected, and diversity of the detection process is achieved.

The invention relates to the technical field of soil health analysis, in particular to a tobacco planting soil health analysis and fertilization prediction system. Comprising a server, a health analysis module and a prediction management module, soil state parameters are obtained through soil microorganism activity action state analysis, soil nutrient state analysis and soil moisture state analysis of tobacco planting soil, then the soil change degree is analyzed, a prediction updating instruction is generated, and comprehensive detection, analysis and evaluation of the soil are achieved. By comprehensively analyzing the state parameters of the soil, the demand change of the soil, including the opportunity and quantification of fertilization, pesticide application and irrigation, can be accurately predicted, so that the growth efficiency and yield of crops can be improved; meanwhile, through the generated prediction notice, a farmer or a manager can be guided to take corresponding measures in time, it is guaranteed that soil and crops grow in the optimal state, and malnutrition, pest and disease damage and other problems are avoided.

The invention discloses a battery cell detection device and method, and belongs to the field of battery cell detection. A battery cell detection device comprises a detection table and an appearance detector fixed to the top of the detection table, the top of the detection table is rotationally connected with a station disc, multiple sets of detection boxes are arranged on the station disc at equal intervals, a driving part for driving the station disc to rotate is arranged in the detection table, and the battery cell detection device further comprises a turnover mechanism. The turnover mechanism is arranged in the detection box; through the arrangement of the station disc, simultaneous operation detection of multiple stations is achieved, the detection effect is effectively improved, battery cells passing through the lower portion of the appearance detector can roll through the cooperation of the attraction effect of the magnetic area and the upper magnetic plate, therefore, comprehensive detection of the battery cells to be detected is achieved, the detection efficiency is effectively improved, and the detection cost is reduced. And by means of self-rolling and self-overturning, the friction effect can be effectively reduced, and the protection effect on the battery cell to be tested is improved.

The invention discloses a self-growing robot and an environment exploration method thereof. The self-growing robot comprises a driving storage mechanism, a plastic film body, a flexible sensing unit, a signal receiving module, pneumatic artificial muscles arranged on the inner wall of the plastic film body and a tail end guiding sensing mechanism. The flexible sensing units are attached to the surface of the plastic film body, the sensing units are designed in an array mode, detection of obstacles in the environment is achieved while flexibility is guaranteed, the tail end guiding sensing mechanism is used for sensing the posture of the robot, information of a proximity sensor, an encoder and a gyroscope is combined, and the detection accuracy of the robot is improved. Navigation in an unknown environment can be realized; besides, pre-bending is set according to environment obstacle information, the environment obstacle is used for guiding the robot to grow, a larger area can be covered by combining pneumatic artificial muscles, a proximity sensor is used for detecting angular points in the environment according to the method, and the environment is explored by adopting a depth-first strategy; and a relatively high environment exploration rate can be realized within relatively few exploration times.

The invention discloses an alloy die casting ultrasonic flaw detection device which comprises an operation table and an ultrasonic flaw detection host, the ultrasonic flaw detection host is fixed on the operation table, a flaw detection assembly is arranged above the operation table, and the flaw detection assembly comprises a movable frame, a mounting frame, an inclined plate, a detection probe, a ball, a cross rod and a second spring. And a mounting frame is arranged on the front side of the movable frame. According to the ultrasonic flaw detection device for the alloy die casting, in the using process of the device, when a movable frame moves downwards to drive an ejector block to move downwards to make contact with a workpiece, a coating roller makes contact with the workpiece at the moment, and the ejector block is stressed to move relative to a fixed frame and cooperates with the effects of a piston, a cylinder, a one-way air inlet valve, a one-way air outlet valve and a guide pipe; according to the coupling agent coating device, air in the cylinder can enter the coupling agent storage barrel, a coupling agent in the coupling agent storage barrel can be discharged to the coating roller under the action of air pressure, and the coating roller rolls through downward movement of the device, so that the coating effect of the coupling agent can be realized, and detection is facilitated.

The invention discloses a brake spring piece flatness detection system and method, and relates to the technical field of automobile part processing. The brake spring piece flatness detection system comprises a data acquisition module, a data analysis module, a data screening module, a dynamic monitoring module and a comprehensive judgment module, wherein the data acquisition module is used for acquiring static detection data of N groups of spring pieces randomly extracted from the same batch; the static detection data comprises top view shooting picture data and side view shooting picture data of N groups of spring pieces, which are acquired at equal intervals by utilizing an optical detection instrument, and size data of the spring pieces, which are acquired by utilizing a laser radar, and the technical key points are as follows: through analysis and processing of the static detection data and the dynamic detection data, the overall quality of the spring pieces is shown; related workers can conveniently and quickly know whether the spring pieces meet the requirements of automobile braking or not, high-quality products can be quickly screened out from numerous spring pieces, use is convenient, secondary detection is not needed in the follow-up process, and good application prospects are achieved.

The invention relates to the technical field of cement detection, and discloses a cement detection device for road construction, which comprises a base, a lifting rotating mechanism and a detection mechanism, a stand column is mounted behind the top of the base, a cross rod is mounted at the top of the front end of the stand column, a control box is mounted at the front end of the cross rod, and a driving motor is arranged in the control box; the bottom of the driving motor is connected with a stirring structure, and a cement storage barrel is arranged above the base; the lifting rotating mechanism is fixedly arranged at the front end of the stand column and is connected to the extension rod and the stirring rod in a penetrating manner; the detection mechanism is connected to the bottom of the lifting rotating mechanism and comprises an impact detection structure and a solidification state detection structure. According to the cement detection device for road construction, comprehensive detection of cement is achieved, impact detection of cement hardness and strength and cement solidification state detection are facilitated, the impact condition in actual construction can be simulated, cement quality can be accurately evaluated, and important reference is provided for construction.

The invention relates to the technical field of stamping part surface detection, in particular to a stamping part surface detection method and a welding slag cleaning device.The detection method is based on three-dimensional point cloud data processing, is not affected by environment light changes, can more accurately separate a defect area from a background, and is high in detection accuracy. The welding slag can be effectively identified even at a complex geometrical shape or a deep concave position, comprehensive detection of a stamping part with a complex shape can be realized, and the accuracy and the efficiency of detection are remarkably improved; by presetting the basic threshold value and dynamically adjusting according to the reflection characteristics, it can be guaranteed that the stamping part can obtain stable detection and cleaning results under the detection conditions of different light environments.

The utility model relates to a wafer edge detection device, comprising a supporting seat which comprises a supporting platform; the buffering unit comprises a supporting frame, a first driving part and a plurality of supporting parts, the supporting frame is arranged on the supporting platform, the supporting parts are arranged on the supporting frame, the first driving part is used for driving the supporting frame to move in the height direction, and a wafer to be detected can be placed on the supporting parts; the rotating unit comprises a base, a clamping assembly, a lifting rod and a second driving part, the base is arranged on the supporting platform, a through hole is formed in the center of the base, the second driving part is used for driving the base to rotate, and the detection camera unit is used for collecting images of the wafer to be detected. According to the utility model, comprehensive detection of the edge of the wafer to be detected can be realized, and the detection precision is effectively improved, thereby ensuring the comprehensiveness and accuracy of the detection result.

The invention provides a social robot detection method based on a KAN graph neural network model, and the method comprises the steps: carrying out the data preprocessing of an identified social account data set, carrying out the adaptive synthesis sampling through employing an ADASYN method, and carrying out the balance processing of the data set; after the balanced data set is input into the model, graph structure data is processed through two layers of convolution, the first layer extracts preliminary features, the second layer further learns multi-level graph features, in each layer of convolution, a KAN linear layer enhances the nonlinear expression ability of the features by using B spline interpolation, and the feature extraction efficiency is improved; the model is helped to better capture various social interaction type relationships in the graph data; neighborhood between nodes is used for learning representation of the nodes, the nodes with higher representation capability can be learned through two convolutional layers for training, classification prediction is further performed through a softmax function, and a final prediction result is obtained, so that social robot detection is realized.

The invention relates to a detection kit for epithelial-mesenchymal transition in EB virus infected thymoma and application, the detection kit comprises a quantum dot labeled antibody E-Cadherin and a quantum dot labeled antibody Vimentin, a quantum dot is coupled with a specific antibody, a thymoma paraffin section sample is dyed through the quantum dot and the specific antibody conjugate, and the EB virus infected thymoma is obtained. The conjugate can be specifically connected with an EMT related antigen expressed in the slice, so that the quantum dots can be used for qualitative or quantitative detection of the EMT-generated thymoma cells. The quantum dots are used for replacing fluorescent dye, direct marking of tissue slices can be achieved, local positioning of epithelial cells and mesenchymal tissues in the thymoma microenvironment is achieved, number detection and position detection can be achieved, and EMT condition evaluation is facilitated.

The invention provides a water turbidity detection method and device and electronic equipment. The method comprises the following steps: acquiring the current flow velocity and temperature of a to-be-detected water body through an ultrasonic meter; determining the measurement amplitude of the ultrasonic signal under the flow velocity and the temperature, and calibrating the measurement amplitude to obtain a calibrated amplitude; determining a current amplitude attenuation rate under the flow velocity and the temperature based on the calibration amplitude and a standard amplitude; and determining the water turbidity of the to-be-detected water body based on the relationship between the turbidity and the amplitude attenuation rate and the current amplitude attenuation rate. According to the water quality turbidity detection method and device and the electronic equipment, the existing structure and circuit design of an existing ultrasonic water meter can be utilized, turbidity detection can be achieved without adding an independent optical turbidity detection sensor, and then the water quality turbidity can be detected by monitoring the envelope amplitude change condition of the ultrasonic receiving signal in real time. And the turbidity change of the water quality is monitored.

The invention relates to the technical field of sheet metal part bending detection, in particular to a method and system for judging sheet metal part bending accuracy based on a three-dimensional model, and the method comprises the steps: determining an error compensation parameter through historical data of a target production line, and calculating a bending simulation adjustment parameter through combining a sheet metal part design drawing, an expected process and a bending simulation parameter; according to the design drawing, the expected process and the bending simulation adjustment parameters, generating a sheet metal part finished product three-dimensional model by utilizing three-dimensional modeling software, and extracting a target expansion view and a target panoramic reference view of the sheet metal part finished product; respectively carrying out bending accuracy detection on the images to obtain an expanded image deviation result and a panoramic reference image deviation result; if the deviation results are qualified, the bending accuracy is qualified; otherwise, the bending simulation parameters and the design drawing are adjusted, and bending simulation adjustment parameter calculation, three-dimensional modeling and bending accuracy detection are repeated until all deviation results are qualified. Therefore, actual production conditions are met, and the bending precision of the sheet metal part is improved.

The invention relates to the technical field of optical detection and image processing, in particular to a mobile phone lens window glass defect detection method. Comprising the following steps: acquiring a high-resolution image of a window glass sample by imaging equipment by adopting a multispectral imaging technology; inputting the acquired image into a pre-trained deep learning model for preprocessing; determining whether defects exist or not according to an output result of the deep learning model, marking the positions and types of the defects, and generating a detection report containing defect geometric parameters; if the defect exists, the severity of the defect is judged according to preset standards and geometric parameters of the defect, and the standards comprise the size, the shape, the depth and the position of the defect; and the detection result and the image data are stored in a cloud database to support subsequent quality analysis and process improvement. By combining the multispectral imaging technology, the deep learning model and the fuzzy logic system, the quality detection level of the window glass is comprehensively improved.

According to the workpiece surface coating bonding strength scratch detection device and method provided by the invention, the workpiece surface characteristics can be identified, so that the movement path of the scriber is planned, the scriber can change the movement direction along with the change of the surface shape of the workpiece, and the detection of the workpiece with the curved surface characteristics can be realized. The scratch detection device comprises a three-axis moving mechanism, a scriber unit, an observation unit and a workpiece clamp, the scriber unit is mounted on the three-axis moving mechanism, can move on the coating surface of the workpiece to be tested along a planned scriber moving path, scribes on the coating surface of the workpiece to be tested by using a scriber, and applies a test load to the workpiece to be tested; the observation unit is installed on the three-axis moving mechanism, can observe the surface shape of a workpiece to be measured and is used for planning the moving path of the scriber. Meanwhile, the observation unit can also be used for observing the appearance of a scratch scratched by the scriber unit on the surface of the coating of the workpiece to be detected; the workpiece clamp is used for clamping a to-be-tested workpiece.

The invention relates to the technical field of deep learning, in particular to a lithium battery fault recognition method and system based on a recurrence plot and an improved AlexNet model, and a storage medium. The method comprises the following steps: acquiring one-dimensional time sequence data from various sensors of the lithium battery; preprocessing the one-dimensional time sequence data to generate original fault waveform data; converting the original fault waveform data into a recurrence plot; the recurrence plot is input into a pre-trained improved AlexNet model for feature extraction, the improved AlexNet model processes feature vectors in the recurrence plot through a multi-head self-attention mechanism, and the processed feature vectors are input into a classification layer for fault type prediction of the battery; and obtaining a battery fault type prediction result output by the classification layer. The objective of the invention is to solve the problem of how to predict possible faults in a lithium battery.

The invention relates to the field of crankshaft detection, in particular to an adjustable support for crankshaft detection and a detection device.The adjustable support comprises an installation guide rail, a supporting mechanism is installed on the installation guide rail through positioning bolts, and the supporting mechanism is used for installing and fixing a crankshaft to be detected so that the crankshaft to be detected can be in a horizontally rotatable state; wherein the detection device further comprises a controller, an alarm, a positioning part and a detection part on the basis of the adjustable support, the positioning part is used for emitting positioning laser, and the detection part is used for detecting a main shaft and a connecting rod shaft of the crankshaft to be detected and feeding back a detection result to the controller. According to the invention, through flexible adjustment of the adjustable support, detection requirements of crankshafts of different specifications and sizes are met, and the problem of adaptability of crankshaft detection is solved; through the detection device capable of being flexibly adjusted, detection of the crankshaft is conveniently and rapidly completed, and the problem of low efficiency of a traditional detection mode is solved.

The invention provides a carbon slide plate loss estimation method, a data processor and an estimation system, and the method comprises the steps: determining a temperature change score of a carbon slide plate at each first set position in response to the obtained first wavelength variation of a temperature grating string; determining a temperature change value of the carbon slide plate based on the temperature change score at the first set position; in response to the obtained second wavelength variable quantity of the strain grating string, determining a stress change value of the carbon slide plate at each second setting position in combination with the temperature change value; based on the stress change value of the carbon slide plate at each strain grating string, determining the wear thickness value of the carbon slide plate at the second setting position of each strain grating string; and determining the actual wear condition of the carbon contact strip based on the wear thickness value of the carbon contact strip at each second set position. Therefore, the temperature grating sensor arranged on the carbon slide plate can be utilized to detect the temperature change value in the carbon slide plate, and further, the actual wear condition of the carbon slide plate can be detected with high precision.

The invention discloses a single-photon laser radar imaging system and method for underwater target detection, and belongs to the technical field of detection imaging. The imaging system comprises an optical system and a control system, and the optical system comprises a transmitting unit, a receiving unit, a single-photon detector and a scanning unit. An optical path of the optical system adopts a transmitting-receiving combined optical path structure, the transmitting unit transmits an optical signal, the scanning unit receives the transmitted optical signal and a control signal and reflects an echo optical signal, the echo optical signal generated by the optical signal transmitted by the transmitting unit is received by the receiving unit, the echo optical signal is converged to the single-photon detector, and the single-photon detector is used for detecting the single-photon. The single-photon detector performs photoelectric conversion; the control system comprises an FPGA control board, and the FPGA control board is connected with a communication module, a gating signal module and a scanning signal module. The system is compact in structure and reasonable in design, detection of weak optical signals can be realized by using a reasonable optical path structure and the single-photon detector, and the farthest detection distance is increased.

The invention discloses a steel rail geometric dimension detection method based on three-dimensional laser scanning. The method comprises the following steps: S1, arranging a three-dimensional laser scanner and a prism ball; s2, establishing a design coordinate system, and registering prism ball point cloud coordinates collected by a three-dimensional laser scanner; s3, steel rail point cloud is collected, and the collected steel rail point cloud is preprocessed; s4, extracting section contour features of the steel rail; and S5, the size of the steel rail is detected and corrected. The three-dimensional laser scanning technology is adopted to detect the steel rail, compared with a traditional contact type measuring method, the measuring speed can be remarkably increased, the labor intensity of workers can be reduced, the continuous high-precision steel rail can be obtained through fitting by collecting a large amount of point cloud data, and the measured data accuracy is higher.

The invention discloses a machine vision detection device and method, and the device comprises a base, the top of the base is provided with at least four groups of clamping structures at equal angles, the lower end of each group of clamping structure is connected with a group of corresponding triggering structures, the triggering structures are arranged on a cavity and the base, and the clamping structures are arranged in the cavity. The number of the cavities corresponds to that of the clamping structures, and the cavities are formed in the base. The trigger structure is connected with a tooth block ring. The machine vision detection device and the method thereof can realize transverse 360-degree multidirectional detection of an object. In the process, the trigger structure plays a key role and can trigger the two groups of clamping parts to perform switching clamping, so that the clamping parts are ensured to clamp the object without shielding in the process that the detection probe detects the object in a surrounding manner, and the interference on the comprehensive detection of the object is avoided.

The invention discloses a quantum sensor based on an optical fiber type diamond NV color center and a preparation method of the quantum sensor. The quantum sensor comprises diamond particles used for generating the diamond NV color center emitting fluorescence; the multimode optical fiber is used for bearing diamond particles and transmitting optical signals; the diamond particles and the multimode optical fiber form a metal microcavity through silver plating on the surfaces, the metal microcavity is used for binding light to circulate and converge in the cavity, the diamond particles and the multimode optical fiber are coupled, the metal Ag film is plated on the surfaces of the diamond particles and the multimode optical fiber to form a metal microcavity structure, the binding light continuously circulates and converges in the cavity, and therefore the binding light can be effectively restrained. The laser excitation efficiency and the fluorescence collection efficiency are effectively improved, the signal intensity detected by the photoelectric detector is enhanced, and high-precision quantum information detection of NV color center spin characteristics and the like is realized, so that ultra-high spatial resolution detection of a physical field is effectively served.

The invention belongs to the technical field of waste gas treatment, particularly relates to a dust removal, desulfurization and decarbonization absorption and desorption device for ship tail gas and a use method of the dust removal, desulfurization and decarbonization absorption and desorption device, and aims to overcome the defects that in the prior art, when gas is treated, the gas cannot be detected and cannot be secondarily treated. A first treatment tank and a second treatment tank are fixedly mounted at the top of the base, the first treatment tank is located on one side of the second treatment tank, a first mounting cover is fixedly mounted on the inner wall of the top of one side of the first treatment tank, one side of the first mounting cover extends to the outer side of the first treatment tank, and one side of the first mounting cover fixedly communicates with an air connecting pipe; after the gas is treated, the treated gas can be detected, and if the treated gas does not reach the emission standard, the gas can be subjected to secondary treatment, so that the quality of the discharged gas reaches the emission standard.

The invention discloses a target bounding box detection system and method based on a millimeter wave radar, and the system comprises a point cloud data generation module which collects a millimeter wave intermediate frequency signal through the millimeter wave radar, and generates the point cloud data of a target; the data optimization module is used for removing noise in the point cloud data by using a reflection saliency coefficient and enhancing the point cloud by using a cumulative matrix; the frame detection module is used for carrying out receiving end beam forming on the millimeter wave intermediate frequency signal to obtain a target reflection energy spatial distribution map, fusing the target reflection energy spatial distribution map and the optimized point cloud data to obtain a fused multichannel feature map, and inputting the fused multichannel feature map into a two-dimensional contour prediction neural network to obtain a two-dimensional contour image of the target; and finally detecting a bounding box of the target in combination with the optimized point cloud data. According to the method, the applicability and reliability of target bounding box detection are improved, and the method can reliably work in various environments with poor light conditions by utilizing millimeter wave radar related algorithm design.