85results about How to "Time consuming to overcome" patented technology

The invention discloses an implement method of parallel fluid calculation based on an entropy lattice Boltzmann model, and provides a parallel implement mode of the entropy lattice Boltzmann model based on the GPU of a mainstream video card nVIDIA in current general calculation territory. The fact that ELBM simulated time is shortened by one third than the simulated time on a CPU is achieved by measuring the speed-up ratio of fluid parallel calculation and comparing the performance index of lattice number updated per second, and applying the CUDA on the GPU of the nVIDIA video card. A method of directly approximately solving a parameter alpha is more effective than an iterative method, namely, and time can be reduced by 31.7% averagely. The implement method of the parallel fluid calculation based on the entropy lattice Boltzmann model can use hardware resources of the system fully, and verifies the parallel calculation mode of the entropy lattice Boltzmann model from the actual operation level, and therefore efficiency of the whole parallel fluid calculation is improved obviously.

The invention relates a method for detecting the quantity of antenna ports in an LTE (Long Term Evolution) system and relates to the technical field of mobile communication. According to the LTE system, the quantity of the antenna ports is hidden in a broadcasting channel through CRC (Cyclic Redundancy Check) mask. The method comprises the following steps of: reading a reference signal under 1 / 2 / 4 antenna port conditions from a received signal, carrying out SNR(Signal Noise Ratio) estimation by using SSS (secondary synchronization signals) and protection non-load carrier, and when the SNR (Signal Noise Ratio) is larger than a threshold value, judging the quantity of antenna ports by using a power ratio of the reference signal of the 1 / 2 / 3 antenna port to a reference signal of a 0th, and otherwise, judging the quantity of the antenna ports by using the reference signal of each antenna port and a related value of a local reference signal.

The invention relates to a sensor communication system and a method for conducting monitoring through the sensor communication system. According to the system and the method, behaviors of livestock in a real cultivation scene can be monitored and analyzed by integrating the internet of things with the computer visual technology, a video surveillance and sensor network software and hardware system based on the wireless WiFi technology is set up for the first time, and the problem that a traditional cable is high in installation cost, long in consumed time and difficult to expand is solved. The behaviors of dairy cattle in the real cultivation scene can be monitored and analyzed by integrating the internet of things with the computer visual technology, monitoring and analysis for multiple broad categories of daily behaviors can be achieved, meanwhile, the occurrence of abnormal events can be monitored, and an alarm can be automatically given for the abnormal behaviors.

The invention discloses a process capable of simultaneously producing and separating high-purity EPA (eicosapentaenoic acid) ethyl ester and high-purity DHA (docosahexaenoic acid) ethyl ester. With crude fish oil as raw material, supercritical fluid extraction, base catalysis and esterification, supercritical counter-current extraction, urea clathration and supercritical rectification are carried out, and EPA ethyl ester and DHA ethyl ester with the purity being more than 90% are obtained respectively. By adopting the process disclosed by the invention, the EPA ethyl ester and the DHA ethyl ester can be effectively purified and separated, and the contents of the obtained EPA ethyl ester and DHA ethyl ester are respectively more than 90%; and no residual solvent exists in the product, preoxidation value is less than 2meq / kg, anisidine value is less than 10, acid value is less than 0.5mgKOH / g, polymer is less than 0.2%, no stearin is contained, and the defects of the prior art that cost is high, process is complex, long time is consumed and safety is difficult to guarantee can be overcome.

The invention discloses a molecular probe for detecting African swine fever virus, a kit and application thereof, so as to solve the technical problems of poor reliability and poor practicability of the existing detection technology. The nucleotide sequence of the molecular probe for detecting African swine fever virus is shown as in SEQ ID NO.2. An RT-PCR reaction solution for detecting African swine fever virus is prepared, lyophilized powder for the RT-PCR reaction solution for detecting African swine fever virus is prepared, an African swine fever virus detection kit is prepared, and an African swine fever virus qualitative and quantitative detection method is provided. The designed molecular probe and the built RT-PCR detection method have the characteristics of high sensitivity, goodrepeatability and strong specificity, the reliability and the usability of ASFV detection by the PCR technology are improved, pollution-free and high-flux detection is realized, and ASFV nucleic acids in susceptible animals and related animal products can be detected.

The invention is a micro-flow control chip-based single-cell endosome analysis, made on a chip platform. First, exert voltage on cell pool and earth cell waste liquor one, the electric power drives the cells to flow from the cell pool to the cell waste liquor one, observe by microscope, and when the cells reach the exit of the detecting channel, make the electrodes of the two pools suspend; at the same time, exert high voltage on run buffer liquor pool and cell cracking liquor pool, and earth waste liquor pool; in this voltage mode, ensure the cells, the cell cracking liquor, screening agent, and run buffer liquor to enter separating channel in order to complete cracking and separating single cell; and detect the endosome on two ends of the channel. It can make single or parallel analysis on the endosome.

The invention provides a three-dimensional modeling method nuclear system combining a dome-screen camera and a depth camera, which relates to the technical field of three-dimensional imaging modeling.The system at least comprises the depth camera and a dome-screen camera. The method comprises eight steps that S1, a dome-screen camera lens and a depth camera lens are fixed at the same position inthe space; S2, photographing is carried out by using the dome-screen camera; S3, preliminary positioning is carried out through a VSSLAM algorithm; S4, the three-dimensional information and space information in the three-dimensional picture are adopted for calibration; S5, further accurate positioning is carried out; S6, a dense point cloud is formed; S7, the accurate dense point cloud is furtherobtained; S8 digital modeling is carried out to generate a three-dimensional model; S9 mapping of the three-dimensional model is automatically carried out according to the color information in the two-dimensional panoramic photo. The invention is advantageous in that the three-dimensional model has the advantages of the dome-screen camera and the depth camera at the same time, and the three-dimensional model is more accurate; the defects that the angle of the depth camera is single, the vision field is limited, and the dome-screen camera is prone to generating errors are solved.

The invention relates to a medical instrument, and especially relates to a multifunctional full-automatic cell and solution treating instrument. The instrument comprises an ultrafilter, a liquid storage container, a peristaltic pump, an electromagnetic valve mixer and a temperature controller, wherein special components in a cell suspension or solution are displaced by a method of instillation combined with ultrafiltration and dialysis, or the cell suspension or solution is diluted and concentrated to desired final volume through accurate control. The multifunctional full-automatic cell and solution treating instrument effectively overcomes the shortages of the existing related treatment methods in the aspects of causing damage of cells, possibly causing cell pollution, taking long time, and low efficiency and the like; and the multifunctional full-automatic cell and solution treating instrument is capable of adding or removing a low-temperature protector quickly and steadily without pollution, and also capable of controlling the volumes and the temperatures of the cell suspension and the solution quite accurately and reducing various cell damages possibly caused in the treatment process. The multifunctional full-automatic cell and solution treating instrument has the advantages of small volume, simple structure, little time consumed and high controllability, and is convenient to operate.

The invention relates to a self-service shopping method. The method comprises the steps: firstly scanning the goods in a user shopping cart by a self-service cash register one by one, obtaining the goods information, and generating a shopping list after the confirmation of a user; secondly carrying out the online payment of the shopping list in a mode of card swiping through a POS machine or in a mode of code scanning through a mobile terminal; thirdly printing a shopping ticket through a bill printer of the self-service cash register; finally enabling a worker to verifying the shopping ticket at an exit of a supermarket, and completing the shopping. The method carries the steps of information scanning through the self-service cash register, the self-service settlement, shopping ticket printing and ticket verification at the exit, can carry out the purchase of commodities in the supermarket quickly and conveniently, keeps the shopping interests of a user in the supermarket, and saves the time of the user in waiting for settlement.

The invention belongs to the technical field of microwave electro-vacuum, in particular to an optimization design method of an electron gun of a traveling wave tube based on a multi-objective optimization algorithm. The method includes the steps of taking geometric structure parameter of the electron gun as the optimization parameter; obtaining the initial structure value of electron gun by the rapid synthesis design method, and determining the approximate range of the optimized parameters; taking the cathode emission current, beam waist radius and electron injection distance of TWT electron gun as the optimization objectives, using a new method based on non-dominated genetic algorithm NSGA-II to approximate the global optimal solution, and finally analyzing the trajectory crossover of allindividual electron beam in the optimal solution set, and selecting the solution with the best laminar flow property as the final design scheme to realize the optimization of TWT electron gun. Underthe premise that the electron gun macroscopic electric parameter cathode current, beam waist radius and range have met the design requirements, the invention considers the microscopic electron trajectory and selects the electron gun structure with the best laminar flow property of the electron beam, further improves the performance of the electron gun and has high efficiency.

The invention, which belongs to the field of environmental monitoring and electrochemical sensors, relates to a tetrabromobisphenol A molecularly-imprinted electrochemical sensor, and a preparation method and application thereof. A glassy carbon electrode modifying a multi-walled carbon nanotube and gold nanoparticles is used as a working electrode; a polymer film of polypyrrole and tetrabromobisphenol A is prepared by using a cyclic voltammetry method; and then tetrabromobisphenol A is eluted to obtain a molecularly-imprinted electrochemical sensor for direct electrochemical detection of tetrabromobisphenol A. Compared with the traditional detection method, the provided method has characteristics of convenient operation, high sensitivity, high specific recognition ability, high reproducibility and strong anti-interference ability.

The invention discloses a skin-whitening skin-tendering freeze-dried facial mask containing adenium obesum essence and a preparation method thereof. The facial mask contains freeze-dried powder. The freeze-dried powder comprises the following components: 0.1-15 parts of adenium obesum leaf cell extract, 0.1-5 parts of hydroxyethylurea, 0.5-5 parts of PEG / PPG / polybutylene glycol-8 / 5 / 3 / glycerol, 0.1-5 parts of glycerol, 0.1-8 parts of sodium hyaluronate, 0.5-5 parts of a natural thickener, 0.1-5 parts of papain, 0.1-3 parts of calcium pantothenate, 0.1-3 parts of urea, 0.1-5 parts of caprylyl glycol, 0.1-5 parts of amino acids, 0.1-5 parts of moisturizing factors, 0.1-5 parts of ethylhexylglycerin and 5-60 parts of water. After being used, the freeze-dried facial mask disclosed by the invention makes the skin healthy, white, fine, smooth and compact, and can delay aging, repair damaged cells and enhance the resistance of the skin with long-time use.

The invention discloses a monte carlo based method for simulating a carbon / carbon composite material texture. The method comprises the steps of performing chemical kinetics modeling for pyrolytic carbon matrix deposition; performing monte carlo modeling for pyrolytic carbon deposition process; compiling C++ program of simulated pyrolytic carbon deposition process; setting the program into 20000 MCS, wherein 1MCS=50*50 cycles; the simulation results are drawn into a map. According to the method, the characteristic of monte carlo simulation randomness is utilized successfully; the deposition of pyrolytic carbon on the matrix surface during CVI process is simulated, so as to really reflect the influence of the CVI process parameters to the whole reaction process; the method is of good technical guide significance for industrial production.

The invention discloses a product technological process compiling method and device, equipment and a storage medium, and the method comprises the steps: obtaining product parameter information which comprises a plurality of sub technological process names; automatically generating a product total process flow by using the product parameter information and a process rule engine model, wherein the process rule engine model comprises a plurality of sub-process flow standard names, sub-process flows and corresponding relationships thereof. The device comprises an acquisition module and a generation module. According to the invention, the product parameter information is obtained;. By using a progress rule engine model, the total process flow of the product is automatically generated, the technical problems that in the prior art, manual writing of process flow parameter information is long in time consumption and high in error rate are solved, automatic writing of the product process flow is achieved, manpower is separated from early-stage writing work, and the manufacturing efficiency of the PCB product in the product design link is greatly improved.

The invention relates to a method for preparing cuprous oxide-indium trioxide solid solution superfine powder. The method comprises the following steps of: performing ultrasonic chemical reduction coprecipitation reaction on 9.99 to 99.99 percent of metal indium (In), copper sulfate (CuSO4.5H2O) and concentrated sulfuric acid (H2SO4) which serve as raw materials by using a sodium hydroxide (NaOH) solution as a precipitating agent, a glucose (C6H12O12) solution as a reducing agent and absolute ethanol (CH3CH2OH) as a stabilizing agent and a dispersing agent; reducing and aging at constant temperature; performing centrifugal separation; washing, and dispersing; and drying, calcining, and thus obtaining the cuprous oxide-indium trioxide solid solution superfine powder. The obtained cuprous oxide-indium trioxide solid solution superfine powder has the particle size of 30 to 50nm, has spherical particles, high purity and a high dispersing property and is distributed uniformly; and the method is simple, has a short production period and is convenient to operate.

The invention relates to the technical field of engineering construction lofting, and specifically discloses an automatic point lofting robot and a method. The automatic point lofting robot comprisesa crawler type trolley and a point alignment module, wherein the point alignment module comprises a 360-degree prism, a minitype prism bar, a minitype range finder, a prism point finding and driving device, an automatic leveling device and a point marking module; the 360-degree prism and the point marking module are connected through the minitype prism bar and are always in coaxial arrangement; the prism point finding and driving device is connected with the upper part of the minitype prism bar and is used for moving the minitype prism bar so as to accurately align to to-be-lofted points; theautomatic leveling device is connected with the upper part of the minitype prism bar and is used for leveling the point alignment module; the minitype range finder is arranged on the minitype prism bar and is used for measuring elevation of the points; the point marking module is arranged at the lowest part of the point alignment module and is used for marking the to-be-lofted points. By adoptingthe automatic point lofting robot disclosed by the invention, automatic lofting of the points can be realized, the efficiency is high, short time is consumed, and the point accuracy is higher.

The invention discloses a design method for a broadband coaxial window, and belongs to the technical field of vacuum electronic device microwave energy transmission windows. According to the method, the coaxial window size preliminarily meeting the requirements is rapidly obtained by the condition of no reflection of center frequency by utilizing an equivalent circuit model of the broadband coaxial window through combination of the microwave network theory, and then structural parameters are scanned and optimized within a small range by utilizing three-dimensional electromagnetic field numerical simulation software so that the structural size meeting the performance requirements can be rapidly obtained. The initial structural size of the broadband coaxial window meeting the specific performance requirements can be rapidly and accurately obtained, problems of high time consumption and high computer resource consumption caused by designing by means of three-dimensional electromagnetic simulation software purely can be overcome, and thus the effective method is provided for the high-speed and high-efficiency design of the high-performance coaxial window.

The invention provides a method for removing silicon thick-film on a silicon carbide substrate, and belongs to the technical field of film deposition. The method employs a chemical reagent to remove silicon thick-film and comprises the steps of: first removing organic pollutants on the surface with an alkali and acid solvent; then using a mixed acid solution to corrode and remove the silicon thick-film; and finally cleaning the surface of the silicon carbide substrate. The invention has the following beneficial effects: the method can quickly remove the silicon thick-film without damage of silicon carbide substrate and surface shape, overcomes the technical risk problems of periodic physical polishing method, such as long period, large time consumption and easy change of silicon carbide substrate mirror embryo surface, effectively improves the processing efficiency of the silicon carbide substrate, and has very high practical value.

The invention provides a rapid manufacturing method of a 3D printing dental orthodontic invisible brace based on the DLP technology. A tooth jaw model comprises a dental crown, gum and a base, wherein the inside of the dental crown, the gum and the base are of solid structures in the spatial structure, and the adopted manufacturing material is a photosensitive resin material used for 3D printing. For a design method of the tooth jaw model, a digital model is designed through a computer, and a processing method of the tooth jaw model is the 3D printing technology of DLP (mask exposure forming), and sequentially comprises the following steps: (1) acquiring and repairing the digital model; (2) carrying out model data stratification; and (3) carrying out 3D printing forming. With the rapid manufacturing method, the surface roughness of the gypsum tooth jaw model prepared by adopting the traditional method is improved, the weight of the tooth jaw model is reduced, the digital model is reserved, so that the storage of data is facilitated, and the production and transportation costs are saved.

The invention discloses a cleaning method of a metal nanowire. The cleaning method comprises the coating procedure, cleaning procedure, dispersion procedure and removing procedure. The cleaning method of the metal nanowire is characterized in that a protective layer is formed on the surface of the metal nanowire, coagulation of the metal nanowire is effectively prevented through the protective layer during follow-up suction filtration washing, a solvent, organic polymers and inorganic salt are removed through suction filtration washing, and finally, the protective layer on the surface of the metal nanowire is removed. The cleaning method has the beneficial effects that the quick and efficient cleaning technology of the metal nanowire overcomes the defects that centrifugal washing time is long, operation is complex and efficiency is low, coagulation of the metal nanowire is effectively prevented through the protective layer during follow-up suction filtration washing, the metal nanowire is quickly and efficiently cleaned, the film forming phenomenon during the filtration washing process is effectively avoided, yield of a product is increased, and quality of the product is improved.

The invention discloses an edge computing user task unloading method in the field of mobile communication. The task unloading problem for a user to a base station in the dense cellular network is notonly influenced by environmental noise, but also influenced by same-channel interference. According to the method, a user task unloading condition when the same-channel interference exists in the heterogeneous network is considered; a communication model is established. An algorithm based on numerical estimation is proposed to estimate the concept of the initial temperature of a simulated annealing algorithm. The approximate initial temperature is searched by adopting variable-step-length numerical estimation. The problem that algorithm time is too long due to the fact that random initial temperature is used is solved, then the simulated annealing algorithm is used for circularly changing a target server of a user, the searching process of a solution is controlled through a cooling function combining logarithms and polynomials, and a task unloading scheme enabling the average unloading time delay of the user to be minimum is searched for. Compared with a traditional method, the methodcan obtain an optimization scheme more quickly, and has better practicability.

The invention discloses a kind of genome chip for analyzing the microbial community structure in an acidic environment. The probe of chip includes the entire genomic DNA of at least ten thiobacillus ferrooxidans, at least two thiobacillus acidophilus sulfur oxides, at least one eosinophils thiobacillus, at least one acidithiobacillus albertensis bacteria, at least six thiobacillus oxidation of ferrous - Lo, at least eight heterotrophic bacteria acidophilus, at least one metal leaf sulfur bacteria, at least one Acidianus tengchongensis, at least a Metallosphaera sedula and at least two of thermophilic bacillus curing.

The invention discloses a method for establishing a pancreatic cancer model in a wild-type adult mouse through lentiviruses. The method comprises lentiviral vector construction, virus packaging, titerdetermination and in-vitro verification, virus injection, and pathological identification. The method comprises selecting an oncogene Kras with the highest variation rate of the human pancreatic cancer and cancer suppressor genes of Tp53, Cdkn2a and Cdkn2b, overexpressing the oncogene KrasG12D in the same lentiviral vectors, simultaneously, knocking down Tp53, Cdkn2a and Cdkn2b through shRNA, carrying out virus injection, and after 40 days, detecting significant pancreatic tumors in the adult mouse which is killed by the pancreatic cancer. The method has a short induction period and high andstable incidence, is easy to operate and fast builds the adult mouse pancreatic cancer model. The pancreatic cancer model simulates the genetic mechanism of human pancreatic cancer to the maximum extent, can be used to study the pathogenesis of human pancreatic cancer and the screening of new therapeutic drugs and therapeutic target points and provides a reference for the treatment on human pancreatic cancer.

The invention discloses a designing method of a millimeter wave varied box type window. The designing method comprises the following steps of: S1, determining a type and a size of a rectangular waveguide and a diameter of a round waveguide according to the scope of an operating frequency of a microwave window; S2, selecting a material, a diameter and a thickness of a medium window plate; S3, determining whether the material, diameter and thickness of the medium window plate, which are selected in the step S2, meet matching conditions by utilizing a microwave network theory; S4, determining a length of the round waveguide by using the microwave network theory; S5, determining whether the transmission characteristic of the millimeter wave varied box type window meet the demand by using the microwave network theory; and S6, carrying out modeling and performance analysis on the determined millimeter wave varied box type window by using three-dimensional electromagnetic simulation software so as to obtain the millimeter wave varied box type window which meets the performance requirements. According to the invention, the millimeter wave varied box type window which meets the specific performance requirements can be accurately and quickly obtained and an effective method is provided for the quick and efficient design of a high-performance microwave energy transmission window.

The invention provides a method of making lotus paste stuffing. The technical scheme includes that the method includes firstly, peeling and coring lotus seeds, namely, cleaning the lotus seeds prior to drying or frying the lotus seeds and peeling and coring, secondly, making the lotus seeds into powder, namely making the dried, peeled and cored dry lotus seeds into powder with a food grinder, a dry grinder or other machines, thirdly, mixing the lotus seed mash, namely putting the lotus seed powder into a pot, adding water weighing 6-15 times of the lotus seeds, evenly mixing the lotus seed powder, heating up to 100 DEG C and stirring while keeping heating, fourthly, obtaining the fished product, namely putting the mixed lotus mash into the pot, adding sugar, stir-frying on and on at medium cooking temperature, adding oil for once to three times, dishing up the lotus paste until the lotus seed mash absorbs the soil completely. The method is energy-saving, timesaving, highly efficient, and applicable to manufacturers making lotus paste stuffing or lotus paste cakes.

The invention relates to the technical field of foundation and foundation engineering, and discloses a deep and large foundation pit wireless intelligent monitoring system which comprises an automatic data acquisition unit used for acquiring foundation pit monitoring data and a wireless remote transmission unit used for wirelessly transmitting the acquired data, and a monitoring service platform which analyzes the foundation pit monitoring data. The deep and large foundation pit monitoring system integrating automatic data acquisition, wireless data transmission, sensor free networking and remote intelligent control is developed by using the Internet of Things technology, real-time, accurate and continuous data are provided for monitoring of the deep and large foundation pit, the deformation trend of the foundation pit is predicted, a scientific early warning mechanism is established, and the safety state of the foundation pit is judged so that safe and smooth construction of the foundation pit is guaranteed, the defects that a conventional manual monitoring means needs more personnel investment, time consumption is long, errors are large and data processing is lagged are overcome, meanwhile, data collection is continuous, the dynamic change conditions under various working conditions can be reflected, and real-time monitoring and early warning are achieved.

The invention relates to a CFD technology-based fluid system and a design method thereof. The CFD technology-based fluid system comprises a CPU module, a GPU module, a flash memory module, an SSD solid-state disk module, a signal input end and a signal output end, wherein the CPU module is connected with the GPU module and the flash memory module separately; the SSD solid-state disk module is connected with the flash memory module; and the signal input end and the signal output end are connected with the flash memory module.

The invention discloses a sulphur dioxide detecting system and a sulphur dioxide detecting method. The system disclosed by the invention comprises a liquid path unit, a gas path unit and a detecting unit, wherein the liquid path unit is used for absorbing and enriching sulphur dioxide gas in atmosphere through an absorbing liquid to react to generate a sulphite solution; mixing the sulphite solution and a detecting reagent, and feeding the mixed solution into the detecting unit; the gas path unit is used for purifying air and feeding the purified air into the detecting unit; the detecting unitis used for detecting a light emitting signal and converting a detecting result into a sulphur dioxide concentration numerical value when the mixed solution of the sulphite solution and the detectingreagent is in contact with the purified air to perform a gas-liquid phase chemiluminescence reaction. The method disclosed by the invention corresponds to the system. According to the system and themethod, the absorption and enrichment as well as the subsequent high-sensitivity detection of the sulphur dioxide in the atmosphere are realized under the premise of ensuring real-time continuous detection; the system and the method have the advantages of simple device, high detecting sensitivity, few interference factors and the like.

The invention discloses a method for determining a chlorine content in a compound fertilizer by using energy dispersion X-ray fluorescence analysis, which belongs to the technical field of analysis detection. The method mainly solves the problems of low excitation efficiency, poor result precision, low sensitivity and low efficiency for direct detection of a chlorine element in the prior art. Themethod is characterized in that a compound fertilizer sample on market is pre-treated, a standard sample and a to-be-measured sample are prepared, an energy dispersion X-ray fluorescence spectrometeris used for measuring a silver element in the sample, a characteristic peak area of the silver element is detected, a scaling curve is substituted, and the percentage content of the silver element inthe sample is obtained, according to a chemical quantity relation of silver and chlorine in silver chloride, the content of the chlorine element in the compound fertilizer can be calculated. The chlorine content in the compound fertilizer can be measured through the indirect measurement method, and the method has the characteristics of accuracy, convenience, economy, and easy operation.

The invention discloses a commodity classification method for a fusion attention graph. The method comprises the steps that a clear and complete commodity front picture is collected as a template image, the template image is extended to obtain a training image, the training image is matched with the template image to obtain an attention graph, RGB three channels are superimposed with the attentiongraph to form four channels, training data is composed, an image classification model is obtained by training, four-channel data of a to-be-classified commodity image is input into the image classification model to obtain predicted classification and a corresponding score of the to-be-classified commodity image. The method can enlarge the influence of fine features of commodity patterns on the classification model, restrain non-significant features of patterns of attention graph to participate in training, and improve the accuracy of the commodity image classification.