122results about How to "Fast and accurate calculation" patented technology

A tunable super-lens (TSL) for nanoscale optical sensing and imaging of bio-molecules and nano-manufacturing utilizes negative-index materials (NIMs) that operate in the visible or near infrared light. The NIMs can create a lens that will perform sub-wavelength imaging, enhanced resolution imaging, or flat lens imaging. This new TSL covers two different operation scales. For short distances between the object and its image, a near-field super-lens (NFSL) can create or enhance images of objects located at distances much less than the wavelength of light. For the far-zone, negative values are necessary for both the permittivity ε a permeability μ. While well-structured periodic meta-materials, which require delicate design and precise fabrication, can be used, metal-dielectric composites are also candidates for NIMs in the optical range. The negative-refraction in the composite films can be made by using frequency-selective photomodification.

The invention discloses an automatic generation method of a cable bridge based on an EXCELVBA (visual basic for applications) technology and computer aided drawing. The method comprises the following steps: a, defining a weight conversion table of specification, cross area and unit length of a cable in a computer system; b, inputting a cable list and a routing table into the computer system; c, calculating the cable cross area capacity and weight of each branch section according to the cable list and the routing table; d, defining the cable cross area of a single-layer bridge to obtain the number of bridge layers of each branch section; and e, generating a material list, a cable bridge diagram, a cable passage trending schematic diagram and a cable network node diagram according to the number of bridge layers in the step d. According to the method, the cable bridge diagram and the material list can be generated automatically, so that cable construction and design checking are facilitated.

The invention discloses a Coriolis mass flow meter cloud transmission digital signal processing device and method. The device comprises a Coriolis mass flow meter, wherein the Coriolis mass flow meter is provided with two magnetoelectricity sensors, a driver and a constant flow source, the two magnetoelectricity sensors transmit collected signals to a differential amplification circuit corresponding to the magnetoelectricity sensors, the differential amplification circuit transmits processed signals to a DSP through an AD sampling circuit corresponding to the differential amplification circuit, and the driver is in communicating connection with the DSP through a feedback type digital driving module. According to the method, digital driving is conducted by means of the feedback type non-linear gain control algorithm fast and stably, frequency is tracked with the Newton LMS algorithm in an adaptive mode timely and accurately with the accuracy capable of reaching 0.01%, and the accuracy of calculation of phase difference achieved by means of the DTFT algorithm with a temperature compensation function can reach the 0.02% industrial level. Therefore, the digital signal processing algorithm is a Coriolis mass flow meter signal processing method with high accuracy and strong timeliness.

The present invention discloses a photon and electron dose calculating method under a magnetic field based on a GPU Monte Carlo algorithm. The method comprises: 1, collecting data; 2, determining the optimal thread count and transport task batches of a GPU; 3, employing the Monte Carlo algorithm to calculate the photon and electron dose of each batch under the magnetic field effect; and 4, performing statistics of the dose results based on the GPU rapid atom addition. The photon and electron dose calculating method under the magnetic field based on the GPU Monte Carlo algorithm can rapidly and accurately calculate the photon and electron radiation dose under the magnetic field effect and can be used for magnetic resonance to guide the dose calculation in the radiotherapy in real time so as to improve the accuracy and the speed of the dose calculation in the radiotherapy and improve the effect of radiation treatment.

The application provides a method, processor, and system for performance parameter calibration of a micro-electro-mechanical-system (MEMS) acceleration sensor. A closed-loop control rotation platform is driven to carry out 360-degree-rotation multi-point positioning on a to-be-measured MEMS acceleration sensor under the field of gravity and an input shaft accelerated speed, an output shaft accelerated speed, a pendulum shaft accelerated sped, and an actual output quantity of the to-be-measured MEMS acceleration sensor are obtained; on the basis of a preset model equation, an expected output quantity is obtained; and least-square-method fitting operation is carried out the expected output value and the actual output value to obtain various model parameters of the to-be-measured MEMS acceleration sensor. The deviation value and scale factor of the to-be-measured MEMS acceleration sensor among the model parameters do not contain quadratic nonlinear coefficients, cubic nonlinear coefficients, cross coupling sensitivity and cross coupling coefficients. Moreover, an influence on the final calibration result precision by the precision limitation of the standard sensor in the prior art is eliminated and thus the calibration precision and the calibration efficiency of the system can be improved substantially.

A tunable super-lens (TSL) for nanoscale optical sensing and imaging of bio-molecules and nano-manufacturing utilizes negative-index materials (NIMs) that operate in the visible or near infrared light. The NIMs can create a lens that will perform sub-wavelength imaging, enhanced resolution imaging, or flat lens imaging. This new TSL covers two different operation scales. For short distances between the object and its image, a near-field super-lens (NFSL) can create or enhance images of objects located at distances much less than the wavelength of light. For the far-zone, negative values are necessary for both the permittivity ∈ a permeability μ. While well-structured periodic meta-materials, which require delicate design and precise fabrication, can be used, metal-dielectric composites are also candidates for NIMs in the optical range. The negative-refraction in the composite films can be made by using frequency-selective photomodification.

The invention discloses an accurate radiotherapy planning system, comprising a CT data input module, a machine data module, a radiotherapy planning database, a target positioning module, a dosage computation module, a computer, an on-line dosage inversion module, a dosage collection module, a beam setting and preset ideal dosage distribution module, a setting beam module, a calculating beam dosage response matrix and determining beam intensity distribution module, a first determination module, a first evaluation result module, a calculating beam generated dosage and optimization module, a second evaluation result module, a second determination module and an output module. The accurate radiotherapy planning system has the characteristics of reasonable design and simple structure, and can accurately and rapidly calculate a target dosage, and substantially increase radiotherapy efficiency; in addition, an evaluation system makes dosages more accurate, is suitable for clinical usage, improves radiotherapy effects and reduces patient treatment pains.

The invention discloses a proton and heavy ion dose calculation method under a magnetic field based on a GPU Monte-Carlo method. The method comprises the following steps: 1) collecting data; 2) determining an optimal thread count and transport task batches of a GPU; 3) calculating proton and heavy ion radiation dose of each batch under the magnetic field by utilizing the Monte-Carlo method; 4) calculating radiation dose of secondary particles of each batch under the magnetic field by utilizing the Monte-Carlo method; and 5) summarizing dose results based on a GPU fast atom addition method. The method can quickly and accurately calculate the proton and heavy ion radiation dose under the action of the magnetic field, and can be used for dose calculation of an MRI real-time guided proton and heavy ion treatment planning system, thereby improving accuracy and speed of dose calculation in the treatment planning system, and improving radiation treatment effect.

The invention provides a GRRM model-based groundwater environment risk evaluation method for a hydrogeological unit. The method comprises the steps of performing risk source assignment according to a risk source type based on spatial distribution of a risk source through a hydrogeological division risk evaluation unit; by taking groundwater quality, groundwater level and groundwater source area sensitivity as risk suffers, calculating risk suffer sensitivity; determining damage degrees of the risk source to a groundwater system and the risk suffers, and obtaining an exposure response coefficient; calculating a groundwater anti-pollution performance index by integrating weighted hydrogeological parameters; and obtaining a groundwater relative risk evaluation result through relative risk model calculation, and performing representation on a groundwater environment in combination with the groundwater relative risk evaluation result. By obtaining the risk source assigned value of the risk evaluation unit, the risk suffer sensitivity, the exposure response coefficient and the groundwater anti-pollution performance index, the groundwater relative risk evaluation result can be comprehensively, quickly and accurately calculated and analyzed, so that the reliability of groundwater relative risk evaluation is improved.

The invention discloses a catering quick counting system based on RFID (Radio Frequency Identification), and belongs to the field of RFID radio-frequency technology. The catering quick counting system comprises an RFID electronic tag, an RFID price marking instrument, an RFID counter and an upper computer, wherein the RFID price marking instrument comprises a display, a keyboard and a microprocessor; the RFID electronic tag is fixedly arranged on a tray main body; the RFID price marking instrument is connected with the RFID electronic tag; information on the RFID electronic tag is read and written through the RFID price marking instrument; the RFID electronic tag, the RFID counter and the upper computer are connected in sequence; the information on the RFID electronic tag is read through the RFID counter; and the read information is transmitted to the upper computer for processing. The system provided by the invention is low in manufacturing cost without influencing dinnerware, and also solves the problems that total manual counting is slow to count and mistakes are easily made, so that the RFID electronic tag is comprehensively utilized, and the system is high in counting efficiency, quick and accurate to count and can be widely applied to the field of catering service.

The invention discloses a system for rapidly determining an accurate imaging scheme for belt material defect detection. The system comprises an accurate calculation resolution testing module, a compound imaging system module, a calibration system module and an image parameter calculation and scheme determination module, wherein the accurate calculation resolution testing module is used for measuring a defect size of a tested sample and determining an imaging resolution; the calibration system module is used for executing rapid focusing before imaging, positioning a focusing position, calibrating a light source of a camera and determining an imaging distance of the camera; the compound imaging system module is used for imaging defects of a motion sample and finishing imaging at a light field, a dark field, a transition field and sidelight according to different light source angles; and the image parameter calculation and scheme determination module is used for estimating and calculating the image resolution, selecting the model of the camera, calculating the imaging distance according to the imaging resolution and the imaging angles, carrying out calibration calculation and image processing, and outputting the final imaging scheme. The system has the beneficial effects that the automation degree is high, the calculation is accurate, rapid calibration and surveying are realized, and the testing efficiency is greatly improved.

The invention relates to a device and system for determining a sequence of blood vessel condition parameters of a blood vessel by utilizing a computer. The device includes a memory, a processor and acomputer executable instruction stored in the memory and capable of running on the processor. The device is used for acquiring an image block sequence of a blood vessel path; and based on the acquiredimage block sequence of the blood vessel path, predicting a sequence of blood vessel condition parameters of the blood vessel path by utilizing a trained deep learning model formed by series connection of a data stream neural network, a recurrent neural network and a conditional random field model. By utilizing advantages of the recurrent neural network in sequence learning and advantages of a convolutional neural network in image learning, and through end-to-end model training, the device overcomes several critical problems in rapid computing of a blood vessel condition parameter sequence, such as computing speed increasing, manual intervention reducing for feature extraction and accuracy ensuring.

The invention discloses a function pass rate estimating method for an integrated circuit, aiming at solving the problem of low accuracy of the existing estimating method. Adopting different pass rate estimating methods for different types of errors, the invention comprises the steps as follows: each layer of plane layout of the to-be-estimated integrated circuits is numbered according to wire mesh; the error shape feature is extracted in production procedures corresponding to of the plane layout of the to-be-estimated integrated circuits; the error shape feature is divided into three grades of a circular outline, an elliptic outline and a random outline; the key areas of the errors of the circular outline, the elliptic outline and the random outline are calculated on a selected plane layout respectively, summing up to gain the total key area of the three errors on the selected plane layout; the pass rate of corresponding processes and the pass rate of the integrated circuit in the selected plane layout are estimated by utilizing a normal method according to the total sum. The invention characterized by high estimating accuracy is applicable for affirmation about the pass rate loss estimation of chips caused by the errors of random shape during the producing process.

The invention discloses a soil porosity detection method based on an SFS (Shape from Shading) algorithm. The detection method comprises the following steps: acquiring a microscopic image of soil to be detected, and recording a light incident angle; selecting M rows and N columns of sampling pixels from the microscopic image of the soil to be detected, obtaining height values of all sampling pixels by utilizing Jacobi iterate algorithm; setting height threshold parameters, deleting the sampling pixel in the M rows and N columns of sampling pixels, wherein the difference of the height value of the deleted sampling pixel and an average height of the periphery adjacent sampling pixel is more than the height threshold, and linearly interpolating the deleted sampling pixel or region according to the height value; removing the height value with the small probability by fitting a normal distribution curve of the sampling pixel, performing height correction to the height value of each sampling pixel; and obtaining the porosity of the soil to be detected according to a porosity calculation formula. According to the soil porosity detection method based on an SFS algorithm, the soil porosity is calculated by recovering a three-dimensional structure of the soil surface; the detection method is more rapid and accurate than other methods; the test cycle of related research can be greatly shortened; and the accuracy rate of related research results can be improved.

The invention relates to a total station instrument and a dial calibration method thereof. The method comprises the following steps: acquiring an image of a simulated star, calculating a first difference between the centroid pixel abscissa and the center pixel abscissa in the image by using a star point extraction algorithm and a second difference between the centroid pixel vertical coordinate andthe center pixel vertical coordinates of the image, fitting a first relationship model of a horizontal dial value and the first difference and a second relationship model of a vertical dial value andthe second difference, and respectively substituting the first difference and the second difference of a star to be detected into the first relationship model and the second relationship model to obtain the horizontal dial value and the vertical dial value of the star to be tested. The total station instrument and the dial calibration method thereof realizes the conversion between pixel coordinates and the dial coordinates of a measured robot, get rid of the limitation of previous obtaining of the dial values human eye aiming observation, and provide a model and data basis for realizing fastand accurate automatic astronomical measurement.

The invention discloses a porosity detection method for soil, relating to the technical fields of agriculture and computer pictures. The porosity detection method for the soil comprises the following steps of: S1, acquiring the microscopic images of a soil sample, and pre-processing the microscopic images, so as to convert the microscopic images to greyscale images and remove noise; S2, generating a greyscale co-occurrence matrix of the greyscale images after the noise removal; S3, calculating the characteristic values such as contrast CON, correlation COR, inverse difference moment IDM and entropy ENT of the greyscale co-occurrence matrix; S4, fitting the contrast CON, the correlation COR, the inverse difference moment IDM and the entropy ENT according to preset fitting indexes to obtain the fitting equation of the porosity; and S5, substituting the contrast CON, the correlation COR, the inverse difference moment IDM and the entropy ENT of the greyscale co-occurrence matrix of the to-be-detected soil in the fitting equation to obtain the porosity of the to-be-detected soil. With the adoption of the porosity detection method for the soil disclosed by the invention, the porosity of the soil can be rapidly and accurately calculated.

The invention discloses a notch part local stress-strain calculation method under high-temperature multi-axis loading. The notch part local stress-strain calculation method comprises the steps of analyzing and determining a notch part local stress-strain state under high-temperature multi-axis loading; Reading seven real stress strains and constitutive model variable parameters in the previous step, and calculating a yield function f1 in the previous step; Calculating nominal stress and strain, and determining loading; Judging whether the current loading is elastic loading or plastic loading by utilizing the yield function in the last step; Calculating a yield function f2 under the current loading step and judging; And storing all real stress strains and constitutive model variable parameters, and returning to carry out next-step solving calculation until the loading is finished. According to an algorithm program prediction result and a finite element simulation result of the method, it is found that the stress-strain peak-valley value is close to the shape of the hysteresis loop. The prediction result shows that the method can be used for well calculating the local stress-strain response of the notched part under the high-temperature multi-axis loading.

The invention discloses a detection method for binocular parallax in a head up display (HUD) system. The method comprises the steps that S1, a double-camera optical module is set, wherein the double-camera module is used for simulating the two eyes of a person; S2, the double-camera optical module is placed behind the HUD system; S3, two cameras in the double-camera optical module are used to shoot test images of the HUD system respectively, and the test images are marked as a first detection image and a second detection image respectively; and S4, the position relation between the first detection image and the second detection image is calculated, and a horizontal parallax angle and a vertical parallax angle are calculated according to the position relation. Through the detection method for the binocular parallax in the HUD system, the binocular parallax can be detected precisely and quickly on an HUD production line.

The invention provides a touch point positioning method of a capacitive touch screen. The touch point positioning method comprises the following steps: detecting capacitance values of all capacitancenodes; calculating capacitance value variation of all the capacitance nodes, and determining a touch area; selecting a peak capacitance node with the maximum capacitance variation in the touch area and a capacitance node group adjacent to the peak capacitance node; and calculating the coordinates of the touch point according to the capacitance value variation and the coordinates of the capacitancenode group or the capacitance value variation and the coordinates of the peak capacitance node and the capacitance node group. The invention further provides a touch point positioning device of the capacitive touch screen. According to the touch point positioning method and device of the capacitive touch screen, the coordinates of the touch point are obtained through calculation by means of the capacitance value variation and the coordinates of the selected capacitance nodes, the touch position can be calculated more accurately and efficiently, and meanwhile various application requirements are met.

An apparatus for calculating insulation resistance according to one embodiment of the present invention comprises: a switching unit comprising a first switch and a second switch which are controlled independently of each other; a first protective resistor and a first reference resistor which are connected in series between a positive terminal and ground of a battery when the first switch is turnedon; a second protective resistor and a second reference resistor which are connected in series between a negative terminal and the ground of the battery when the second switch is turned on; a voltagemeasuring unit configured to measure a first detection voltage applied to the first reference resistor, and a second detection voltage applied to the second reference resistor; and a processor configured to determine whether a measurement error has occurred in measured values of the first detection voltage and measured values of the second detection voltage.

A livable environment system based on artificial intelligence comprises collection modules, a temperature control module and a processing module. The collection modules comprise the temperature collection module, the wind speed collection module, the humidity collection module and the video collection module. The temperature control module is used for adjusting the temperature and the output windspeed according to the calculation result of the processing module. Compared with the prior art, the livable environment system has the beneficial effects that the optimum PMV value is obtained through clustering training, the temperature control module can control the temperature more accurately, and the temperature is suitable for human bodies; the current temperature and wind speed suitable forthe human bodies are calculated through the cuckoo algorithm, the number of iterations is small, convergence is rapid, and the current temperature and wind speed suitable for the human bodies can becalculated and obtained more rapidly and accurately; and each frame in video is collected and analyzed, clothes recognition and posture recognition are carried out through a machine learning method, and the current human body clothes condition and motion state are obtained accurately.

The invention provides a method and a system for calculating parameters of flow heat transfer of a refrigerant in a tube, a medium, a terminal and a battery pack. The method comprises the following steps: setting the pressure in a state equation of a thermodynamic system as an independent variable; calculating a saturated liquid phase Prandtl number vector; calculating heat transfer coefficient vectors of all nodes based on the saturated liquid phase Prandtl number vector; based on an energy relationship, calculating a first relation curve of a coordinate vector and a saturation pressure vector of a node according to the heat transfer coefficient vector of the node and a heat flow loaded on the node; calculating a second relation curve of the coordinate vector and the saturation pressure vector of the node based on a pressure relation; determining the coordinate parameters of the node according to an intersection of the first relation curve and the second relation curve; calculating the heat transfer coefficient at the node according to the coordinate parameters of all nodes. After the method and the system for calculating the parameters of the flow heat transfer of the refrigerantin the tube, the medium, the terminal and the battery pack are adopted, the parameters of the flow heat transfer of the refrigerant in the tube can be rapidly and accurately calculated.

The invention relates to a synchronous degradation process and device for COD (Chemical Oxygen Demand) and N (Nitrogen) of high concentration wastewater in a same pool. Two aerating devices are arranged at the bottom of a biochemical pool which is divided into a first cell aerating area and a second cell aerating area with different dissolved oxygen control by a functional separating wall. Wastewater and a return liquid are mixed and aerated at the water inlet of the first cell aerating area, the mixed liquid at the tail end of the first cell aerating area flows into the second cell aerating area and is aerated, and degradation of COD and N is carried out in the first and second cell aerating areas. A part of mixed liquid at the tail end of the second cell aerating area returns to the water inlet of the first cell aerating area while the other part flows into a cement separating device. Cleaned water after cement separation is discharged. The dissolved oxygen control range of the first cell aerating area is 0.1-0.5mg / L, the dissolved oxygen control range of the second cell aerating area according to property of wastewater is 0.1-1.8mg / L. The invention has the benefits that the COD removal efficiency of the biological synchronous denitrification process is improved, synchronous degradation of COD and N in the same pool is realized, and the process can be applied to domestic wastewater treatment. Therefore, the invention is more applied to high concentration industrial wastewater treatment with COD and N.

The invention provides a pointer type instrument information identification method and device and electronic equipment. For any to-be-identified pointer instrument which needs to be read, the electronic equipment acquires an instrument image of the pointer type instrument; feature extraction is carried out to obtain the position information of a plurality of specific key points including the instrument range of the pointer instrument to be identified, an instrument range starting point, an instrument range ending point, an instrument pointer fixing point and an instrument pointer index point;then, no matter what instrument type the pointer instrument to be identified is, the current reading of the pointer instrument to be identified can be quickly and accurately calculated by directly utilizing the respective position information of the plurality of key points and the instrument range, a specific identification algorithm does not need to be deployed in advance, the method can be suitable for various pointer instruments, and the identification efficiency is improved.

The invention relates to an efficient overall planning method for routing inspection paths and tasks of underground safety inspection personnel of a coal mine, and the method comprises the steps: constructing an efficient overall planning system for routing inspection: based on the Internet, consisting of a safety inspection system platform, a personnel positioning module, a server, an early warning module, a mobile client APP, a database, and information acquisition points; Uploading the information collected by each area of the coal mine to a database, and providing data for calculation of aserver; Data arranged by a server and a mathematical model compiled in a coal mine security check system are applied, various algorithms are adopted, the number of security check persons and inspection routes arranged on all parts of a mine are calculated, and then the number and inspection routes are uploaded to a mobile client; And according to the personnel positioning module, the position andthe inspection information of the security inspection personnel are mastered in real time, and an early warning module can be used for reporting and warning when an emergency occurs. According to theinvention, the method can achieve the quick and accurate calculation of the inspection path planning of the security inspection personnel, achieves the positioning and early warning, improves the safety of a coal mine, improves the working efficiency of personnel, and reduces the enterprise cost.

The application discloses an on-line analysis method for real-time monitoring of the dose of a radiotherapy patient, comprising the following steps: step 1, a first set of detectors on a head of a connected radiotherapy machine collect the beaming dose data of the radiotherapy machine during each radiotherapy treatment in real time, and transmits the data to a microcontroller in real time, and then the microcontroller transmits the data to a special database in a memory for storage; step two, through a second set of detectors arranged on a human body surface, the real-time radiation dose actually received by the human body surface is sent to the microcontroller through a network, and then passed to the special database in the memory via the microcontroller; and step three, by calling the two sets of data in the special database, the microcontroller uses a comparative analysis module for real-time comparative analysis and calculates the dose error, and through continuous iteration of historical data in the special database, the model is optimized once the error is calculated, so that a recommended dose calculated the next time is more optimized.

The invention discloses a multidimensional simulating method for hydrogen retention and desorption in tungsten in a radiation environment. The method comprises the following steps of by means of a binary collision approximation method, calculating a primary off-position damage generated by irradiation and spatial distribution thereof; by means of a density functional theory method, calculating atom scale physical parameters of radiation defects and hydrogen in the tungsten, and estimating the physical parameter of a relatively large defect cluster; and by means of an object dynamics Monte-Carlo method, simulating long-time cooperative evolution of the hydrogen and the radiation defect in the radiation environment. According to the multidimensional simulating method, a DFT, a BCA and an OKMC method are combined, thereby realizing sequence multidimensional analysis of defect evolution. Compared with an existing simulating method, the multidimensional simulating method has advantages of accurately describing hydrogen-defect interaction, improving simulated time and space dimension to above hour / micrometer magnitude, and conveniently examining the influence of micro parameters such asirradiation ion energy, flux and temperature to hydrogen retention / desorption. The multidimensional simulating method is particularly suitable for long-time simulation of hydrogen retention and desorption in the tungsten in the radiation environment.

The present invention provides a full-automatic, high-efficiency and high-accuracy method and device for calculating purity and chromosome ploidy of a cancer sample. Through a hierarchical Gaussian mixture model provided by the invention, the purity and the chromosome ploidy of the cancer sample are rapidly and accurately calculated; the time and the economic cost of purity estimation are saved; the accuracy of calculation results is increased at the same time; the method and the device have broad application prospects in the calculation of the purity and the chromosome ploidy of the cancer sample.

The invention discloses a CTLE value traversal optimization method and a CTLE value traversal optimization device, which are achieved based on BERTscope. The CTLE value traversal optimization method comprises the steps of reading a currently set CTLE value via CSII software, and computing the bit error rate of a receiver by using the currently set CTLE value; judging whether the bit error rate accords with a preset requirement, if no, modifying the currently set CTLE value by using the CSII software, thus obtaining the currently set CTLE value, and computing the bit error rate of the receiver by using the currently set CTLE value until the bit error rate accords with the preset requirement. According to the CTLE value traversal optimization method, the test of the bit error rate of the receiver is achieved based on BERTscope, and meanwhile the CTLE value is modified via the CSII software, thus the problem that in the prior art the optimization of the CTLE value needs to consume a lot of time and is relatively low in work efficiency is solved.

The invention discloses a vehicle video recognition-based hybrid main line toll station traffic conflict evaluation method. The method includes the following steps that: 1, cameras are arranged the toll station square of a toll station to acquire the driving video of vehicles in the toll station square; 2, vehicle identification and tracking are performed in the acquired driving video of the vehicles, the position coordinates of the vehicles in each frame of image are obtained; 3, the data of M frames of images of the vehicles are utilized to calculate the speeds Vi and deceleration indexes DRi of the vehicles in each frame of image; if a vehicle-following conflict occurs, the time index TTCk of a vehicle-following model is calculated, and if a lane change conflict occurs, the time index Tpet of a vehicle steering model is calculated; and 5, peering processing is performed on the DRi, TTCk and Tpet, so that the DRi, TTCk and Tpet can be unified into traffic conflict evaluation indexes STC, and the severity degree of the traffic conflict is judged according to the values of the traffic conflict evaluation indexes STC. With the method adopted, the severity degree of the traffic conflict of the toll station can be quantified. The method has a great application value in decision making such as toll station road safety evaluation, management operation and road traffic flow guidance.