149results about How to "Resolve distortion" patented technology

The invention discloses a corrugated pipe inner wall multi-type flaw detection method based on deep learning, and the method comprises the following steps: S1, collecting an initial corrugated pipe inner wall image and corrugated pipe inner wall images in different illumination environments, and generating an original image; s2, preprocessing the original image to generate a preprocessed image; s3, extracting surface geometric overall features of the inner wall of the corrugated pipe from the preprocessed image through a pre-trained ResNet101 network; s4, inputting the preprocessed image and the surface geometric overall features into an improved StyleGAN3 network to generate an enhanced image; s5, inputting the enhanced image into the multi-scale adaptive deep network, and outputting a defect detection result; and S6, performing post-processing on the flaw detection result, and outputting a final flaw detection result. According to the method, deep learning and geometric characteristic embedding technologies are adopted, so that the defects of the inner wall of the corrugated pipe are accurately detected, and the method has the advantages of high precision, high robustness and comprehensive adaptability.

The invention discloses an intelligent driving environment sensing method and system based on multi-modal data fusion, and relates to the technical field of automatic driving, and the method comprises the steps: firstly obtaining the position information of a vehicle through a GPS, and starting an auxiliary positioning mechanism if the GPS signal is weak; multi-modal data acquisition and preprocessing: after an auxiliary positioning mechanism is started, a vehicle acquires surrounding environment information through an RGB camera, a laser radar and an IMU sensor; performing feature extraction on the RGB image by using an improved deep convolutional neural network, and generating a high-precision 3D bounding box through a region suggestion network; and multi-modal information is fused through a constructed efficient image reasoning framework. Based on multi-modal data fusion and an environment perception result, the method not only can carry out accurate target identification on the surrounding environment in the automatic driving process, but also can carry out global and local path planning.

The invention provides a DEM super-resolution reconstruction method and system fusing gradient feature Transform, and belongs to the technical field of image processing, the reconstruction method is composed of a plurality of residual feature fusion blocks, a multi-scale pyramid segmentation Transform layer in the residual feature fusion blocks fuses multi-head self-attention and pyramid segmentation attention, and the multi-scale pyramid segmentation Transform layer fuses the multi-head self-attention and the pyramid segmentation attention. In order to make the reconstructed DEM closer to reality, a double-filtering convolution block is used, high and low frequency features of the DEM are fully extracted through a parallel structure of two different filtering blocks, and finally super-resolution reconstruction is carried out under double constraints of a gradient domain and a height domain. The method adapts to topographic features of different scales and complexities, and solves the problem that DEM details cannot be accurately recovered by a traditional algorithm.

The invention relates to the field of artificial intelligence and image processing, in particular to a dynamic portrait verification and check result generation system based on machine learning, which comprises multi-modal data acquisition, recursive feedback and dynamic feature optimization, feature completion and correction, and spatial-temporal feature verification and correlation analysis. Performing time sequence optimization through a recurrent neural network, complementing missing features through a generative adversarial network, and correcting time-space consistency by using Bayesian reasoning; the system is combined with a Markov chain Monte Carlo method to dynamically adjust the weight of the multi-modal data, and more accurate time sequence characteristic data is generated; the space-time relevance of the feature data is analyzed through the graph neural network, a final verification result is generated, the identity verification accuracy and consistency in the dynamic scene are ensured, and the method is particularly suitable for passenger verification and security systems in high-flow dynamic scenes such as civil aviation.

The invention relates to a power data federated learning privacy protection method based on efficient homomorphic encryption. The method comprises the following steps: step 1, a client initializes a neural network model; step 2, the server side encrypts and updates local neural network model gradient parameters uploaded by the client side; 3, the server side executes gradient average aggregation operation, the server side does not need to decrypt, the encrypted gradient parameters uploaded by the client side are directly calculated, an average value is solved, and therefore global gradient parameters are updated; step 4, after each client receives the updated global gradient, carrying out decryption calculation operation; 5, loading the updated weight parameters of the neural network model into a local neural network model, and performing neural network model training by using a local data set; the problems that an existing encryption technology is large in operand, large in computing resource overhead and low in algorithm efficiency, and privacy leakage exists in the transmission and analysis process of power data can be solved.

The invention belongs to the field of computational explosion mechanics, and particularly relates to a rapid calculation method for a shock wave and detonation wave propagation process in a non-uniform medium, and the rapid calculation method comprises the following steps: 1, initializing a calculation program; step 2, estimating and calculating step length in each time step; step 3, carrying out integration on the ordinary differential equation by adopting an explicit three-order TVD Runge-Kutta format; 4, solving a Mach number through an A-M relation; 5, data correction is carried out, wherein the data correction comprises regularization and smoothing processing of grids; and step 6, checking whether termination time is reached or not, if so, ending all processes to complete operation, and if not, returning to the step 2 to continue to circularly solve. According to the rapid calculation method, rapid numerical calculation of propagation of shock waves and detonation waves in a non-uniform medium is achieved, wherein the phenomena comprise refraction, transmission, Mach reflection and the like.

The invention discloses a central air-conditioning system indoor thermal environment prediction method based on data driving, which relates to the field of energy conservation and comprises the following steps: screening indoor thermal environment influence factor variables of a building, and performing data preprocessing on the thermal environment influence factor variables; optimizing the BP neural network by using an improved grey wolf optimization algorithm to obtain a data-driven central air-conditioning system indoor thermal environment prediction model; the preprocessed thermal environment influence factor variables are input into a data-driven central air conditioning system indoor thermal environment prediction model, and the central air conditioning system indoor thermal environment is predicted; and performing performance evaluation on the data-driven central air-conditioning system indoor thermal environment prediction model by using the evaluation index to complete data-driven central air-conditioning system indoor thermal environment prediction. The problems that an existing method is low in prediction precision, and a prediction model is prone to being affected by equipment operation data are solved.

The invention relates to the technical field of tire simulation design, in particular to a tire force transmissibility finite element simulation method and system and a program product. According to the method, an axial symmetry finite element model comprising a tire, a rim, internal air and an additional deformable object is established, material attributes and contact pairs are defined, and a three-dimensional tire model is generated in combination with inflation simulation. An air unit with a displacement freedom degree and a self-adaptive grid division technology are adopted. The method comprises the following steps: simulating the embedding, inflating and loading processes of a tire under different working conditions by dynamically switching a contact pair relationship, performing steady-state dynamics analysis on the force transmissibility of the tire, and calculating an acceleration frequency response function and a force transmissibility curve of a tread and a rim center. According to the method, the analysis efficiency is greatly improved, the simulation time is shortened by 80%, meanwhile, a high-precision simulation result is kept, and the method is suitable for force transmissibility evaluation of tires with multiple air pressures and multiple specifications and has wide application value.

The invention discloses an underwater image enhancement method and device based on color perception fusion attention and background CD-ROM separation contrast learning. The method comprises the following steps: extracting color information in an underwater original image based on a designed color extraction module; inputting the color information into a local color recovery branch to obtain reasonably distributed color weight information; performing triple attention branching on the underwater original image to recover details and fuzzy information in the image, and obtaining an underwater enhanced image with rich detail texture information; and inputting the reasonably distributed underwater enhanced image with rich color weight information and detail texture information into a global light recovery branch, adjusting the global light characteristics of the underwater image by adopting a contrast learning strategy, and finally obtaining a recovered underwater image enhancement result. According to the method, the enhancement performance in the underwater scene can be effectively improved, the degradation influence of the underwater complex environment on the image can be eliminated, and the color and texture details of the image can be truly restored.

The invention relates to an ultra-large laser power detection system and method based on vortex light interference of torsion balance measurement, and belongs to the technical field of optical precision testing, and the system comprises a laser to be tested, a reflector A, a fast reflecting mirror A, a reflector B, a reflector C, a torsion balance and a micro-displacement measurement system based on vortex light self-conjugation interference, the reflecting mirror C and the reflecting mirror B are symmetrically distributed relative to the torsional balance; the micro-displacement measurement system based on vortex light self-conjugation interference is positioned below the torsion balance and is used for reversely deducing the power of the laser to be measured through angle rotation of a petal-shaped interference pattern of vortex light; measurement of ultra-large laser power is converted into displacement, then micro displacement is transmitted through the torsion balance and then is subjected to self-conjugation interference of vortex light, the micro displacement can be obtained through the rotation angle of a petal-shaped interference pattern, and therefore the laser power value is obtained. The device has the advantages of high feasibility, novel structure, real-time emergent beam monitoring, high measurement precision and the like.

The invention relates to the technical field of tire dynamics, in particular to a finite element simulation method, device, medium and program for lateral deviation and longitudinal slip composite slip characteristics of a tire. The invention provides a tire lateral deviation and longitudinal slip composite slip characteristic finite element simulation method which comprises the steps of finite element pretreatment, two-dimensional inflation and three-dimensional loading modeling analysis, steady-state simulation analysis, dynamic lateral deviation and longitudinal slip composite slip characteristic simulation modeling analysis and simulation result extraction, the method is high in simulation precision, and meanwhile the simulation efficiency is considered.

The invention discloses a color adjustment method and device for a color electronic paper image, terminal equipment and a computer readable storage medium. The method comprises the steps of obtaining a to-be-processed display image of electronic paper; obtaining a color adjustment instruction; performing corresponding color adjustment on the to-be-processed display image according to the color adjustment instruction to obtain a pre-adjusted display image; a Floyd-Steinberg algorithm is adopted to carry out jitter processing on each pixel point in the pre-adjusted display image, and the updated pixel value of each pixel point is obtained; and for each pixel point, according to the updated pixel value and a preset color mapping table in the electronic paper, calculating a color distance to obtain a color distance set, and taking a color corresponding to a minimum value in the color distance set as a display color of the pixel point. By implementing the method and the device, the problems of limited color expression range and distortion when the electronic paper displays the image can be solved, and the visual experience of a user is improved.

The invention discloses a flexible material surface wettability patterning treatment method and application thereof. The flexible material surface wettability patterning treatment method comprises the following steps that S1, the surface of a hydrophilic flexible substrate is coated with a patterned polyvinyl alcohol aqueous solution, a patterned area covered with the polyvinyl alcohol aqueous solution and an exposed area not covered with the polyvinyl alcohol aqueous solution are formed, and a flexible substrate I in an intermediate state is obtained; s2, a hydrophobic material layer is deposited in the exposed area, a hydrophobic area is formed, and a flexible substrate II in an intermediate state is obtained; and S3, removing the polyvinyl alcohol aqueous solution on the surface of the flexible substrate through washing, and forming a hydrophilic region on the surface of the substrate to obtain the flexible substrate with the wettability patterned surface. The treatment method has the characteristics of soft treatment means and difficulty in damaging the flexible substrate, and the flexible material subjected to surface wetting patterning treatment can be used for preparing a patterned organic semiconductor material and has a good application prospect.

The invention relates to the field of sensor communication, in particular to an intelligent sensor communication method. The method comprises the following steps: acquiring a physical quantity signal in an environment by using an intelligent sensor, generating a corresponding time domain signal, and converting the time domain signal into a frequency domain signal; based on the frequency domain signal, using a dynamic frequency hopping and power consumption perception optimization algorithm to select an optimal frequency; based on the optimal frequency, performing data block coding by using a nonlinear multi-level frequency coding algorithm; and evaluating the path state and the channel performance by using a path optimization selection algorithm, and calculating to obtain the optimal transmission path of each data block. The problems of frequency band overlapping and interference in sensor data acquisition are solved; the problems of fluctuation and inconsistency of sensor signals caused by changes of environmental conditions, reduction of signal reliability due to interference among different frequency channels, signal distortion due to frequency band interference among different data blocks, and fluctuation of channel quality and bandwidth of different paths along with time are solved.

The invention provides a wall interactive projection method and system based on CIM digital twinning, and is applied to the field of projection data processing. The environment change data of the wall projection area is collected in real time, the projection content can be dynamically adjusted in combination with the CIM model, the stability of the projection content in the physical space is ensured, meanwhile, the interaction content and the projection content input by the user are precisely synchronized by dividing the projection subareas and compensating the interaction influence in real time, and the user experience is improved. Projection distortion or jumping is automatically corrected under the environment change, accurate positioning and dynamic calibration are performed by using three-dimensional space coordinate data, and a projection path and angle are optimized, so that the distortion problem caused by angle change or shielding is avoided, and the stability of the projection effect and the user experience are improved.

The invention discloses a fish school identification and classification method in a complex environment based on a particle swarm and improved YOLOv6. The method comprises the following steps: firstly, carrying out image enhancement processing on an acquired fuzzy light image by applying a multi-scale Retinex algorithm on an image acquisition level; secondly, an improved BC-YOLOv6 algorithm is applied to the image recognition level, Backbone of YOLOv6 is combined with BiFormer of a Transform layer of a bidirectional structure, a coordinate attention mechanism module is introduced into the architecture of the BC-YOLOv6 algorithm, the recognition rate of small targets is increased, non-maximum suppression is improved, and correlation functions are designed; and finally, defining a feature matrix and improving a particle swarm algorithm on an image classification level, and obtaining the selected feature matrix to realize image classification. Simulation results show that manpower and computing power cost can be effectively reduced, enhancement processing on the images is remarkably improved, and the monitoring rate, the accuracy, the feature expression ability and the image classification performance of underwater fish school image recognition and classification under the shielding condition are further improved.

The invention provides a display device and a display method thereof. The display device comprises a flexible display screen; the coiling mechanism is connected with one end of the flexible display screen and used for coiling the flexible display screen; the controller is used for generating a first control instruction; the driving mechanism is connected with the controller and the curling mechanism and used for driving the curling mechanism to rotate according to the first control instruction so as to wind or unwind the flexible display screen; the detection part is used for detecting the unfolding degree information of the flexible display screen; and the processor is connected with the detection component and is used for adjusting the display image of the flexible display screen into a target display image according to the unfolding degree information of the flexible display screen. Therefore, it can be guaranteed that the flexible display screen can display the target display image in different unfolding states, the display image can be optimized and adjusted according to the actual visible area, the clearness and authenticity of the display image are kept, and the visual experience of a user is improved.

The embodiment of the invention provides a federated learning method based on a node selection strategy, electronic equipment and a storage medium, and relates to the technical field of federated learning. The method comprises the following steps: determining that client nodes in the federated learning system are not totally selected; obtaining an evaluation value corresponding to the client node, weighting the number of local training samples of the client node by taking a model of the evaluation value as a weight, and obtaining a selected probability of the client node according to the proportion of the number of the weighted local training samples in the sum of the number of the weighted local training samples in the federated learning system; selecting a plurality of client nodes according to the selected probability of each client node to obtain a training set of this round; and issuing model parameters to the client nodes in the training set in the round. According to the embodiment provided by the invention, the overall training time of the federated learning model is shortened in a mode of randomly selecting the nodes.

The embodiment of the present application provides a method for evaluating fine cracks in cement-based materials based on digital image processing. The technical key points are: stretching a target specimen of a high-toughness cement-based composite material ECC, collecting a crack image once at a fixed interval through a camera, and obtaining a background-corrected crack image by performing signal averaging conversion, median filter filtering and other operations on the crack image; obtaining qualified binary images that accurately contain information on fine cracks and wide cracks in sequence through an edge recognition algorithm or a dual threshold algorithm; denoising and enhancing each of the qualified binary images in reverse and forward order according to the time sequence axis of shooting, obtaining the pixel area and boundary coordinates of each crack region through each of the denoised images, identifying the crack width as the projection length of the crack in the loading direction, and obtaining the average width of the crack by dividing the pixel area of ​​the crack region by the crack length, accumulating the crack length and dividing it by the entire width of the specimen to obtain the number of cracks.

The invention relates to a PM2.5 concentration prediction method and system, and the method employs a geometric algebraic space-time diagram recurrent neural network model to predict the PM2.5 concentration of a target region, and the geometric algebraic space-time diagram recurrent neural network model comprises a geometric algebraic-graph convolutional neural network and a geometric algebraic-long and short-term memory network. Comprising the following steps of collecting multi-source data and performing preprocessing; extracting spatial features from the preprocessed data by using a geometric algebraic-graph convolutional neural network; on the basis of the spatial characteristics, utilizing a geometric algebraic-long and short-term memory network to obtain spatial and temporal characteristics of PM2.5; the PM2.5 concentration is predicted according to the spatial and temporal characteristics; the system is used for implementing the method. Compared with the prior art, the method has the advantages of high PM2.5 concentration prediction precision and the like.

The invention relates to a switch cabinet arc detection system based on fluorescent optical fibers, and aims to improve the efficiency and accuracy of switch cabinet arc monitoring in an electric power system. The system is characterized in that the system comprises a host module, a photoelectric conversion module and an optical fiber sensing detection unit, simultaneous detection of multiple switch cabinet arcs can be realized, the detection time is less than 2.5 ms, and the sensitivity and reliability of arc detection are significantly improved. Besides, the system also provides a rapid electric quantity protection algorithm, and a high-reliability arc light rapid protection device with protection action time of 3-5ms is developed by combining voltage, current and arc light comprehensive criteria, so that the problem of rapid clearing of arc light discharge faults is effectively solved. The arc detection system has the advantages of being low in cost, easy to maintain, high in environmental adaptability and the like, and is suitable for arc monitoring in the high-voltage switch cabinet.

The present invention belongs to the field of elevators, and specifically relates to a back-to-back guide rail installation tool for a backpack-type villa elevator, including a working platform substrate, a car guide rail installation assembly, and a counterweight guide rail installation assembly. The working platform substrate has at least one car guide rail installation position and at least one counterweight guide rail installation position. The car guide rail installation assembly is arranged on the car guide rail installation position, which is used to realize the detachable installation of the working platform substrate and the car guide rail and to calibrate the car guide rail. The counterweight guide rail installation assembly is arranged on the counterweight guide rail installation position, which is used to realize the detachable installation of the working platform substrate and the counterweight guide rail and to calibrate the counterweight guide rail. The present invention cleverly utilizes the relationship between the counterweight guide rail and the car guide rail to use a fixed and overall displacement method to calibrate the elevator guide rail. When the counterweight guide rail is calibrated, the car guide rail is naturally vertical, which reduces the installation workload and work intensity.

The invention relates to the technical field of power system fault detection, in particular to a factory power distribution system fault interval section determining method and system and terminal device.The factory power distribution system fault interval section determining method comprises the steps that when a power failure fault of a power distribution system is monitored, voltage and current data within a preset duration before the fault is obtained; comparing the phase current amplitude with a rated current threshold value, judging that the phase current amplitude is an external fault if the phase current amplitude does not exceed the limit, calculating a zero-sequence current if the phase current amplitude exceeds the limit, and comparing the zero-sequence current with a zero-sequence threshold value to distinguish a ground fault or an inter-phase fault; adaptively calling a phase difference analysis process according to an identification result, calculating a phase difference between a zero-sequence current and a zero-sequence voltage during a grounding fault, calculating a phase difference between a fault phase voltage and a fault phase current during an inter-phase fault, and judging whether a fault interval is internal or external according to a phase difference interval; and a structured message is automatically generated by integrating fault types, interval judgment and triggering criterion paths, so that quick and accurate fault positioning and handling are realized.

The invention discloses a reverse time migration imaging method based on solving an adaptive eikonal equation, which relates to the field of exploration geophysical technology, including: S1, inputting a model and source parameters; S2, obtaining a first arrival traveltime reference solution; S3, inputting a forward extension of the source, and saving a forward wave field corresponding to the first arrival traveltime; S4, inputting a reverse extension of a seismic record, and saving a reverse wave field corresponding to the first arrival traveltime; S5, applying imaging conditions for imaging. The method alternately scans in each direction based on the Fermat principle, adaptively selects local plane waves, spherical waves, and refraction wave operators, accurately simulates the first arrival traveltime of near, medium, and far migration, and obtains a first arrival traveltime reference solution; then the traveltime reference solution is used as the excitation moment, the corresponding forward wavefield and reverse wavefield are saved, and then the imaging conditions are applied to obtain a high-fidelity migration image, while also maintaining the advantage of efficient calculation.

The invention discloses a tailoring device with a limiting structure for textile fabric processing, and relates to the technical field of textile fabric processing.The tailoring device comprises a main body, a limiting assembly, a power source, a processor, heat dissipation holes and an operation panel, the limiting assembly is connected with the processor through a signal line, and the operation panel is connected with the processor through a signal line; a limiting assembly is installed on the upper side of the outer wall of the body, a power source is installed on the lower side of the inner wall of the body, a processor is installed on the lower side of the inner wall of the body, heat dissipation holes are formed in the two sides of the outer wall of the body, and an operation panel is installed on the left side of the outer wall of the body. According to the textile fabric tailoring device, by installing the limiting assembly, the function of moving along with a textile fabric in the tailoring process of the textile fabric and limiting is achieved, the problem of distortion in the tailoring process of the textile fabric is solved, deformation in the tailoring process of the textile fabric can be prevented, the yield of tailoring processing of the textile fabric is increased, and the tailoring processing efficiency of the textile fabric is improved. And the subsequent manual processing cost is reduced, and the economic effect is improved.

The application discloses a double-reference parameter fusion DSQM non-cooperative signal separation method, device, equipment and medium, and relates to the technical field of digital signal processing.The method comprises the following steps: receiving an input signal and performing symbol extension; estimating an original input noise standard deviation, a weak signal and a signal-to-noise ratio, and calculating a right shift number and a dynamic weighted fusion coefficient; performing perturbation injection and phase compensation on the extended signal to generate an actual separation input signal, and estimating a noise standard deviation thereof; calculating an adaptive shift number k of the modified signal-to-noise ratio according to the original input weak signal and the actual separation input signal by using a dynamic weighted fusion algorithm; further calculating a left shift number m; amplifying the actual separation input signal by shifting the actual separation input signal left by m bits; sequentially performing right shift n-bit and left shift n-bit to extract a strong signal component; performing time domain subtraction to separate a weak signal component; respectively shifting m bits to the right to recover the magnitude; then respectively performing symmetric saturation truncation and asymmetric truncation to recover the bit width output. The application improves the signal separation quality.

The invention relates to the technical field of substrate glass kiln equipment, in particular to a large-extraction-amount kiln side breast wall supporting device and a mounting method thereof. The large-extraction-amount kiln side breast wall supporting device comprises a side breast wall supporting iron assembly and at least one supporting assembly, the side breast wall supporting iron assembly comprises at least two sections of side breast wall supporting iron, the side breast wall supporting iron is arranged in the length direction of a side breast wall, and a gap is formed between any two adjacent sections of side breast wall supporting iron; the supporting assembly is installed on a stand column in the kiln and used for supporting the side breast wall supporting iron assembly and fixedly connecting the side breast wall supporting iron assembly with the stand column in the kiln. According to the large-extraction-amount kiln side breast wall supporting device, the problem that an original integral supporting iron is distorted and deformed due to expansion in the process of heating up a kiln is solved, and therefore the safety of a kiln side breast wall and a crown in the whole process of heating up the kiln is guaranteed.

The utility model discloses an explosion-proof oil measuring tape, which relates to the technical field of explosion-proof oil measuring tape equipment and comprises a fixed frame, a first connecting block and a second connecting block are respectively and fixedly mounted on the inner side of the fixed frame, a handle is fixedly mounted on the outer side of the second connecting block, and an adjusting structure matched with a limiting structure for use is arranged on the outer side of the first connecting block. According to the utility model, through mutual cooperation among the fixed block, the movable block and the first spring, the flexible rule body is clamped between the pulleys, the flexible rule body can be limited, the flexible rule body can be assisted to be retracted and released, and through the performance of the spring, the pulleys and the movable block can move in the fixed block, so that the flexible rule body is prevented from being damaged. By means of the structure, the measuring tape body can be kept in a flat state all the time to be folded and unfolded, so that the technical problems that in the rolling process, a tape band may be twisted or curled, and damage of the tape band directly affects the service life and the measuring precision of the measuring tape are solved, and the practicability and the service life of the whole anti-explosion oil dipstick are prolonged.

The present application relates to a kind of low-carbon park integrated energy system multi-agent collaborative carbon responsibility allocation method, steps include: the multi-source data of park is collected, with the minimum integrated energy system total operating cost as goal to carry out multi-energy flow optimization scheduling;With the dynamic correction model of two-way carbon flow considering the carbon flow transfer caused by energy storage charging and discharging process, dynamic carbon flow distribution matrix is constructed;Using load carbon responsibility decoupling allocation model, total load is decoupled into multiple sub-load categories, and different carbon responsibility allocation coefficients are given for each type of sub-load;With each interest party in park as the participant of game, Nash bargaining as collaborative optimization framework, maximum all participant utility gain as objective function, Shapley value as the basis for carbon responsibility allocation, multi-agent collaborative optimization is carried out, and the carbon responsibility allocation scheme of each subject is obtained in combination with user carbon responsibility allocation coefficient.Compared with prior art, the present application ensures the fairness, efficiency and operability of carbon responsibility allocation scheme.

The invention discloses a mine safety production monitoring method based on a multi-view optical system, and relates to the technical field of image processing, and the method comprises the steps: obtaining an original mine image through the multi-view optical system; obtaining internal parameters and external parameters of the multi-view optical system as parameter information of the multi-view optical system; acquiring a mine image by using the original mine image according to the parameter information of the multi-view optical system; and performing pipeline splicing processing according to the mine image to obtain a mine safety production monitoring result. The method not only solves the problems of camera distortion and monitoring dead angles in a traditional monitoring scheme, but also improves the quality and real-time performance of the monitored image through an intelligent image processing technology and NPU rectangular data processing, and has remarkable technical advantages and application prospects. And the safety and the working efficiency of mine production can be greatly improved by adopting a multi-view optical system.