125results about How to "Improve the detection rate" patented technology

The invention relates to the technical field of nondestructive testing of constructional engineering, and discloses a laser measuring instrument for nondestructive testing of concrete cavities, which comprises a laser interference measuring assembly, a paraxial sound wave excitation assembly, coaxially arranged anti-interference equipment and a cooperative control processing system. The sound wave excitation assembly performs wide-area excitation on the concrete surface, and the laser interference measurement assembly receives returned light through a clean light path constructed by the anti-interference equipment. The cooperative control processing system establishes acousto-optic time-frequency association by using an adaptive control module, generates a phase compensation angle according to ranging data to eliminate phase lock loss caused by sound wave propagation delay, and monitors a spectral gradient to automatically trigger fixed-frequency locking to capture resonance response. And the signal analysis module uses the phase compensation correction signal, and outputs cavity features through dual data cleaning and thermodynamic diagram analysis. According to the invention, the problems of time-space synchronization and weak signal extraction of long-distance non-contact detection in a complex environment are effectively solved.

The invention discloses a low-altitude unmanned aerial vehicle countering monitoring system based on state feature decoding, relates to the technical field of unmanned aerial vehicle prevention and control, and is used for solving the problem of poor low-altitude target detection and countering. According to the invention, through closed-loop management and control of heterogeneous radar networking detection, signal level fusion processing, threat assessment and resource scheduling and intelligent countering decision, accurate identification and directional countering of the low-altitude unmanned aerial vehicle are realized; according to the method, the detection precision and the micro-Doppler feature extraction capability of a low-speed small target are improved by combining a virtual aperture synthesis technology, then unbalanced resource allocation is implemented based on a threat level, the system efficiency is optimized, and finally seamless migration of an interfered node task is realized by constructing a signal interference situation map. And continuous tracking and effect evaluation on the target are maintained in the active radio frequency countering process, so that the overall efficiency and reliability of low-altitude security and protection in a complex environment are effectively improved.

The application provides a SAR target detection method combining feature fusion and position relationship awareness, comprising the following steps: constructing a real-time target detection model based on a Transformer, introducing a spatial detail compensation path into a mixed encoder and containing a mixed representation enhancement module; a Transformer decoder comprises a position relationship awareness module; inputting a SAR image into a trained target detection network, inputting multi-scale fusion features into the Transformer decoder, adjusting self-attention weight distribution according to the position relationship between queries in the same layer by using the position relationship awareness module, guiding spatial layout reasoning, and outputting a target detection result; and the target detection network is trained by using a weighted intersection over union zoom loss function based on a dynamic IoU weight guiding mechanism as a classification loss function. In this way, the target detection rate, positioning accuracy and overall robustness in a small target and dense scene are improved, and feature confusion under strong scattering interference is alleviated.

The invention relates to a circuit board defect detection method, device and equipment based on multiple sub-models and a storage medium, and is applied to the field of defect detection, and the method comprises the steps: obtaining an original image of a to-be-detected circuit board, aligning the original image according to a preset mother layout, and setting the aligned original image as a to-be-detected panoramic image; segmenting the panoramic image to be detected into a plurality of unit detection images according to a preset segmentation standard; matching a special sub-model from a preset detection model library according to the unit detection image, and performing defect detection on the unit detection image by adopting the special sub-model; and integrating the defect detection results of the unit detection images, and outputting the defect detection result of the circuit board to be detected. The circuit board defect detection method has the technical effect of improving the efficiency and accuracy of circuit board defect detection.

The application provides a child skull fracture automatic detection method and device based on generated data augmentation. The method comprises the following steps: acquiring a first child brain CT image set, and performing pixel labeling and preprocessing to obtain fracture data, suture data and normal data; obtaining a trained fracture lesion generation model and a suture generation model according to the fracture data, the suture data and the normal data; randomly sampling Gaussian noise for multiple times, inputting the Gaussian noise into the fracture lesion generation model and the suture generation model to obtain generated fracture CT patch data and generated suture CT patch data; training a candidate fracture detection model according to the fracture data, the suture data, the normal data, the generated fracture CT patch data and the generated suture CT patch data, and outputting a fracture detection model; and acquiring a second child brain CT image, and detecting the second child brain CT image according to the fracture detection model. The fracture lesion data and the suture data are augmented, so that the resolution of fractures and sutures can be improved.

The application discloses a kind of based on data analysis's abnormal metering identification method and system of petrol pump, it is related to intelligent metering abnormal identification technical field, including, to flow velocity time series respectively extract macroscopic morphological abnormal feature and flow velocity fluctuation feature, to flow velocity difference sequence extract transaction microscopic feature;The application is cooperated modeling by flow velocity time series and flow velocity difference sequence, combines the macroscopic morphological similarity degree matching of abnormal template, flow velocity fluctuation uncertainty interval extension and flow velocity fluctuation visual angle divergence analysis, realizes the systematic characterization to petrol pump metering process;By combining multi-feature discretization modeling and entropy and mutual information analysis mechanism of three-dimensional joint probability distribution, realize the deep mining of statistical dependence relationship and cooperative change mode between multidimensional features, and combined with symmetric measure and skew-symmetric measure, construct two-dimensional decision space, realize the hierarchical identification and dynamic decision of different complexity, different intensity abnormal mode.

The application provides a kind of shipboard aviation equipment composite material damage D-S evidence theory nondestructive evaluation method, belongs to the field of shipboard aviation equipment nondestructive evaluation, it includes the following steps: at time, the triaxial damage area characteristics and triaxial vibration frequency characteristics of the composite material structure of shipboard aviation equipment are synchronously acquired and stored;The triaxial damage area characteristics and triaxial vibration frequency characteristics at time are compared respectively, and the corresponding triaxial damage area change degree and triaxial vibration frequency change degree are obtained;The identification framework of D-S evidence theory is established, and the triaxial mass function of the composite material structure is calculated respectively;According to the created D-S damage judgment model, the classification judgment and early warning of the composite material damage of shipboard aviation equipment are carried out.The application can realize the real-time classification evaluation of the dynamic damage of the composite material component of shipboard aviation equipment, and provide data support for the condition-based maintenance of shipboard aviation equipment.

The invention relates to a tumor-related DNA methylation marker combination, which is selected from at least one of the following: a target sequence as shown in SEQ ID NO: 1-SEQ ID NO: 10, a nucleic acid fragment with the length of at least 15 oligonucleotides or a complementary nucleic acid fragment thereof, and the nucleic acid fragment comprises at least one methylation site indicated by CG. The invention also relates to a detection kit for the methylation marker combination. According to the invention, the combination of the co-methylation states of a plurality of specific methylation regions is used for distinguishing and analyzing the occurrence of tumors, especially kidney cancer, the specific methylation combination has high sensitivity for distinguishing the occurrence of kidney cancer, and the detection method is simple, convenient and feasible. The primer pair combination of the kit overcomes the defect of false positive caused by detection mismatch of a single methylation site in the aspect of primer sequence design, and the interaction between a plurality of methylation biomarker primers and probe pair combinations is considered.

The invention provides a preparation method of a tin-doped scaly nano gallium oxide thin film solar blind ultraviolet photoelectric detector. The preparation method comprises the following steps: carrying out cleaning pretreatment on a single crystal GaN substrate; depositing an Au thin film on the pretreated single crystal GaN substrate to serve as a catalyst layer; the preparation method comprises the following steps: mixing Ga2O3, diamond and SnO2 powder according to a certain mass ratio to prepare a precursor; putting the single crystal GaN substrate deposited with the Au thin film and a precursor into a CVD furnace, and growing according to preset gas flow, pressure intensity, reaction temperature and reaction time to obtain a tin-doped scaly nano gallium oxide thin film; and respectively preparing electrodes on the surface of the tin-doped scaly nano gallium oxide thin film and the surface of the GaN substrate by adopting a radio frequency magnetron sputtering method. The conductivity of the thin film is improved through tin doping, the specific surface area and light absorption are increased through the scaly nanostructure, high-performance deep ultraviolet photoelectric detection is achieved, and the technical problem that in the prior art, the dark current of a device is high is solved.

The application provides application of dimethyl carbonate as a chronic obstructive pulmonary disease diagnostic marker and belongs to the technical field of biomedical detection. The application first finds that the detection rate of dimethyl carbonate in exhaled air is significantly related to chronic obstructive pulmonary disease, and the detection rate of dimethyl carbonate in exhaled air of a chronic obstructive pulmonary disease patient is significantly higher than that of a healthy person. The ROC curve analysis result shows that dimethyl carbonate has extremely high specificity as a biomarker for diagnosing chronic obstructive pulmonary disease, and therefore can be used as a detection target for the diagnosis of chronic obstructive pulmonary disease patients, and has good practical application value.

This utility model relates to the field of biomedical technology and discloses a professional shaker for gynecological liquid-based cytopathology, comprising a base, an outer shell, an inner shell, and a support assembly. A vibrating base plate is mounted on the top of the base, and a vibrating assembly is mounted on the bottom of the vibrating base plate, located within the inner cavity of the base. The outer shell is mounted on the top surface of the vibrating base plate, and a top cover is detachably connected to the top of the outer shell. A first magnetic connection part is mounted on the bottom of the inner side wall of the outer shell. The inner shell is located within the inner cavity of the outer shell, and a second magnetic connection part is mounted on the bottom of the outer side wall of the inner shell, magnetically connected to the first magnetic connection part. The top of the inner shell contacts the top cover. The support assembly is mounted on the inner wall of the inner shell, and several cell fluid tubes are detachably connected to the support assembly. This utility model enables precise vibration and temperature control, allowing for the simultaneous and efficient processing of multiple gynecological samples, thus improving work efficiency.

The invention provides an image processing method, conference equipment and a computer readable storage medium, which can be applied to the technical field of image processing. The image processing method is applied to the conference equipment, the conference equipment comprises a first camera and at least one second camera, and the method comprises the following steps: obtaining a first image based on the first camera, and obtaining a second image based on the second camera; a target image is displayed, the target image comprises at least one object, and the at least one object is obtained by fusing the detection result of the first image and the detection result of the second image.

The invention relates to a binary function clone detection method for graph disturbance and temperature parameter adjustment, and belongs to the field of artificial intelligence safety. The method comprises the following steps: firstly, constructing a function control flow diagram, and generating an attribute control flow diagram for node additional semantic features; secondly, applying structure and instruction disturbance to the attribute control flow diagram to generate diversified positive samples; then calculating function information amount by using structure and grammar features to adjust a temperature parameter of a loss function, and training a graph encoder by using comparative learning; and finally, generating a graph embedding vector of the function pair to be detected by using a graph encoder, and judging whether the function is a cloned function or not through a cosine distance threshold between the vectors. Aiming at the problem that an existing method is difficult to detect functions with few control branches and high grammar repeatability, the method utilizes function information amount to regulate and control contrast learning temperature parameters to increase function functional area indexing, and detection false alarms are reduced; for the problem that overfitting is easily caused by a fixed training data compiling mode, the sample diversity is increased by applying disturbance, and the detection accuracy is improved.

The application discloses a kind of fusion distributed optical fiber acoustics embankment underwater pipe heaving detection positioning system, it is related to water conservancy engineering safety monitoring technical field, comprising: voiceprint signal perception preliminary screening module, for collecting original acoustic vibration signal in strong noise environment, and carry out preliminary purification and separation to it, obtain the intrinsic mode function of different frequency bands, separate out weak pipe heaving suspected signal submerged in background noise.The application maps the seepage network reasoned into high-precision fluid-structure coupling model by introducing digital twin technology, simulates the evolution process under extreme working conditions such as sudden change of water level in virtual space, so that the system can calculate the final breach form, size and position, and the consequences of pipe heaving disaster are pushed from the qualitative judgment to the quantitative prediction, realizing the dynamic deduction and quantitative prediction of pipe heaving disaster process based on physical mechanism.

This invention relates to a membrane tension compensation method based on the detection of uniformity in the thickness of water-soluble membranes produced by dishwashing pods. Addressing the problem that membrane thickness distribution is sensitive to nonlinear tension disturbances under high-speed operation conditions during water-soluble membrane production, and that traditional thickness control methods struggle to achieve precise closed-loop compensation, this invention proposes a method that continuously collects cross-sectional thickness information of the membrane material, constructs a time-space thickness sequence, extracts thickness gradient topological features, combines a multidimensional tension sensitivity feature vector with a dynamic thickness gradient topological graph, and uses a layer-by-layer sparsely connected lightweight graph convolutional network to infer the tension compensation amount. After median filtering and denoising, the result drives the actuator to adjust in real time, achieving high-precision closed-loop compensation for the uniformity of the water-soluble membrane thickness. This method improves the sensitivity and stability of thickness control under nonlinear disturbance environments, while also possessing both real-time performance and computational efficiency.

A crack detection sensor utilizes ultrasonic waves to detect cracks in a test object. It includes a main unit, a transducer, and a coupling layer. The transducer is communicatively connected to the main unit, and the coupling layer abuts against the detection face of the transducer and is used to remove gas from the detection face. This invention, by setting a coupling layer between the transducer detection face and the test object, effectively removes gas between them, ensuring efficient transmission of the ultrasonic signal and thus improving the sensitivity and accuracy of ultrasonic detection.

The invention discloses a method for detecting small defects on the surface of lightweight steel, and belongs to the technical field of industrial machine vision. The method is based on an improved YOLOv8n architecture and cooperatively works through three core technical means: firstly, a Swin Transform module is introduced into a backbone network to model long-distance spatial dependence and suppress complex background interference; secondly, a high-resolution P2 detection branch is constructed, shallow details and up-sampling semantic features are fused, and microdefect characterization is enhanced through bidirectional refining circulation; finally, P2 exclusive adaptive threshold focus loss ATFL is adopted, threshold update is only limited to P2 detection branch samples, and precise optimization of difficult and tiny defects is achieved in cooperation with a gradient directional return mechanism. According to the scheme, the detection recall rate and the positioning precision of the tiny defects are effectively improved while the model parameter quantity is remarkably reduced, and high-precision and light-weight industrial deployment is realized.

This application discloses a testing method for nickel-based superalloys, belonging to the field of material purity testing and evaluation. The testing method includes sample preparation: processing the nickel-based superalloy raw material into cylindrical tensile specimens with a gauge length diameter of 6–15 mm, and polishing the surface; room temperature tensile testing: stretching the tensile specimen at room temperature until it fractures; fracture surface analysis: after fracture, observing the fracture surface using an optical microscope and / or scanning electron microscope, measuring the size of non-metallic foreign matter on the fracture surface, and analyzing its composition to complete the detection. This application has the following advantages: high sensitivity: it can detect micro-non-metallic foreign matter with a maximum Feret diameter of less than 10 μm; high efficiency: the fracture surface can be obtained through room temperature tensile testing, eliminating the need for a high-temperature environment and complex fatigue testing; the segmented tensile process can significantly shorten the testing cycle; for a specimen with a working section of approximately 50 mm, the test can be completed in only about 3 hours.

The invention relates to the technical field of machine vision detection, in particular to an inner package detection and recognition method based on vision detection, which comprises the following steps: synchronously acquiring a multi-view image sequence of a to-be-detected piece through an annularly arranged acquisition device; independently analyzing each image, and generating an initial defect area mask and a local feature descriptor; a three-dimensional defect space distribution model is constructed by fusing multi-view masks, and geometric topology and depth information of the three-dimensional defect space distribution model are extracted to form a comprehensive feature vector. And performing cross-modal correlation analysis on the local features and the three-dimensional comprehensive features to generate an enhanced defect feature spectrum. And calling a pre-training classification model to analyze the atlas, outputting an inner package type, a defect type and a severity level, and generating a quality evaluation report and a processing instruction according to the inner package type, the defect type and the severity level. According to the method, three-dimensional space accurate positioning and multi-feature deep fusion of the defects are realized, and the recognition accuracy and evaluation comprehensiveness of the complex defects are improved.

The application provides a tobacco data cleaning method and a storage medium. The tobacco data cleaning method comprises the following steps: S1, acquiring a to-be-cleaned tobacco data set; S2, performing visual analysis and detection processing on the to-be-cleaned tobacco data set to obtain a first detection result; S3, according to the attribute type of the to-be-cleaned tobacco data set, selecting a corresponding algorithm in a preset abnormal value detection algorithm set to detect abnormal data in the to-be-cleaned tobacco data set, and obtaining a second detection result; and S4, cleaning the to-be-cleaned tobacco data set according to the first detection result and the second detection result to obtain a target tobacco data set. Through the parallel joint detection of multiple algorithms, the accuracy and robustness of abnormal identification are significantly improved, global abnormality, skew distribution abnormality and relative deviation abnormality are fully covered, and the technical defects of high false judgment rate and high missed detection rate of traditional single detection algorithm are effectively solved.

This invention relates to the field of jujube sorting technology, specifically to a high-speed, non-destructive jujube sorting system with a flexible material handling structure and its working method. The system includes a feeding inlet, a feeding conveyor belt, sorting rollers, variable-pitch spiral guide ribs, a full-field transparent detection chamber, a finished product collection bin, and a foreign matter collection bin. The feeding inlet feeds jujubes into the sorting rollers via the feeding conveyor belt, arranging them in single-grain intervals. The full-field transparent detection chamber on the CoreXY planar motion mechanism detects and sorts the jujubes into different finished product collection bins. Diseased fruits and foreign objects such as leaves and branches enter the foreign matter collection bin. This invention achieves efficient, non-destructive, and accurate jujube grading through a closed loop of "flexible upper material handling—full-field perception—non-destructive jujube grading—destruction and compression of diseased fruits and foreign objects."

The utility model discloses a magnetic resonance auxiliary shimming eyeshade which comprises an eyeshade frame, balls are filled in the eyeshade frame, two sides of the eyeshade frame are respectively and rotatably connected with glasses legs through rotating shafts, the interiors of the glasses legs are hollow, and first arc-shaped limiting blocks are uniformly arranged at the bottoms of the glasses legs at intervals. Second arc-shaped limiting blocks are evenly arranged at the top ends of the glasses legs at intervals, first arc-shaped abutting rods are arranged between the first arc-shaped limiting blocks in a sliding mode, second arc-shaped abutting rods are arranged between the second arc-shaped limiting blocks in a sliding mode, and elastic blocks are arranged between the first arc-shaped abutting rods and the second arc-shaped abutting rods in an abutting mode; the first arc-shaped abutting rod and the second arc-shaped abutting rod penetrate through the glasses legs and are connected with extension glasses legs. The small glass beads with the diameter of 2 mm are filled in the eyeshade, the eyeshade has good physical stability and chemical inertness and is easy to clean, disinfect and store for a long time, the glass beads serve as auxiliary shimming devices, the magnetic resonance imaging quality of eye sockets can be improved conveniently, and practicability is high.

The application discloses a PCB defect detection method and device, computer equipment and a storage medium. The method comprises the following steps: collecting resistance data of multiple test points in a to-be-detected area of a PCB to form a resistance data set; sorting the resistance data set and determining the median of the resistance data set as a data median Med; calculating the absolute deviation of each data in the resistance data set from the data median Med to form an absolute deviation set; sorting the absolute deviation set and determining the median of the absolute deviation set as an absolute deviation median MAD; constructing a robust detection threshold for defect determination; and comparing each resistance data in the resistance data set with the robust detection threshold to obtain a PCB defect detection result. The method of the application can realize PCB resistance defect detection with high anti-interference, high detection rate and low false alarm rate, and improve the quality control level of the PCB manufacturing process.

This invention relates to the field of printing machinery manufacturing technology and discloses a printing roller rounding machine with machine vision flaw detection. The machine includes a base with a groove on its top. A movable seat is slidably connected to the inner side of the groove. A flaw detection mechanism is located on the top right side of the movable seat. The flaw detection mechanism includes a bracket, which is fixedly connected to the rear end of the top right side of the movable seat. A hydraulic rod is fixedly connected to the inner side of the bracket, and a connecting rod is rotatably connected to the top of the hydraulic rod. An arc-shaped plate is rotatably connected to the front side of the bracket. This invention uses a motor to drive a lead screw to rotate, which in turn moves the movable seat, enabling the flaw detection mechanism and the rounding mechanism to move freely along the printing roller axis. It can perform full-coverage scanning of printing rollers of different lengths and specifications, and uses a polarizing filter in conjunction with a lamp inside a hollow plate to supplement lighting, eliminating specular high-gloss reflections on the surface of the metal printing roller.

The present application relates to the technical field of signal transmission, and in particular to a smart connected vehicle information security evaluation method, system and device. According to the time interval between adjacent message receiving time stamps in the associated time period at the current time, the timer remaining duration of each message receiving time stamp is obtained. According to the timer remaining duration of each message receiving time stamp, the timeout approximation degree of each message receiving time stamp is obtained. The shape deviation degree of the timeout approximation degree sequence and the preset attack mode sequence is determined. According to the timeout approximation degree and the shape deviation degree, the attack determination index at the current time is obtained. From the attack determination index, the smart connected vehicle information security evaluation result is obtained. The present application can effectively distinguish natural jitter and malicious attacks in a complex mobile network environment, significantly reduce the false positive rate while maintaining a high detection rate for hidden attacks.

The application relates to the field of industrial detection, and provides a flaw detection method, an electronic device and a storage medium. The flaw detection method comprises the following steps: acquiring a to-be-detected image; under the condition that the to-be-detected image is not scaled, the to-be-detected image is cropped to obtain multiple sub-images and position information of each sub-image in the to-be-detected image, and the resolution of the sub-image is the same as that of the to-be-detected image; each sub-image is subjected to flaw detection to obtain a flaw position in the sub-image; and the flaw position in the to-be-detected image is determined according to the flaw position in the sub-image and the position information of the sub-image in the to-be-detected image, so that the problem of flaw pixel loss caused by the reduction of a large image is avoided, and the detection rate of small flaws is effectively improved.

The invention belongs to the technical field of biology, and discloses a plateau dairy cow liquid phase breeding chip and application thereof, and detection objects of the liquid phase chip comprise 75939 SNP sites positioned on a common cow reference genome AR-UCD 1.2. The liquid chip provided by the invention can ensure comprehensive coverage and genetic information representativeness of bovine genomes, and provides a reliable basis for subsequent molecular identification, population structure analysis and molecular breeding application. The gene not only has population genetic polymorphism, but also covers a gene functional region and a non-functional region. Practicality and representativeness of the chip are guaranteed, stability and functional correlation of genetic markers can be considered, and a reliable tool is provided for molecular breeding and genetic improvement.

The application discloses a kind of dynamic detection method and system of malicious code based on double space embedding, applied to network security field, including obtaining and analyzing program running behavior event, extraction program interface identification and calling parameter;Interface identification is mapped to Euclidean space to obtain the first characteristic vector, calling parameter is mapped to hyperbolic space to obtain the second characteristic vector, and the comprehensive characteristic vector of behavior event is formed by fusion;According to the similarity degree of adjacent behavior event comprehensive characteristic vector, dynamically filter out multiple target behavior event sets associated with potential attack behavior;Based on each event set, the corresponding dynamic topology sequence is constructed, and the sequence is analyzed and processed, and the stage characteristic representation of each component element is extracted;According to stage characteristic representation, maliciousness determination and result output are completed.The dynamic detection method of malicious code based on double space embedding provided by the application can accurately detect multi-stage, high-concealment malicious code such as advanced persistent threat.

This invention discloses an underwater target detection device and operating method based on forward-looking sonar partitioning, relating to the field of underwater acoustic signal processing technology. This invention, by displaying the measurement range, divides the beam domain into grids from two dimensions: beam and range, forming multiple local detection grids. It compresses and images the corresponding beam domain data to form image domain data, and determines the detection threshold for each local detection grid based on the signal-to-noise ratio of the corresponding image domain data to identify underwater targets. By dividing the beam domain data of the detection area into grids according to beam and range, this invention can reasonably divide the beam domain data generated by underwater targets, underwater reverberation, and acoustic ray changes in the beam domain data corresponding to the signal domain. It compresses and images the beam domain data of each local detection grid, and accurately identifies the parameters corresponding to underwater targets, underwater reverberation, and acoustic ray changes from the perspective of obtaining the detection threshold from the signal-to-noise ratio of the image domain data, thus accurately identifying underwater targets.

The invention relates to the technical field of machine vision and image processing, and discloses a non-ferrous metal intelligent impurity removal analysis method based on machine vision, and the method comprises the steps: collecting and correcting the hyperspectral time sequence data of the melt surface; constructing a spectrum prior weighted optical flow energy functional to calculate an accurate motion field capable of retaining a material boundary; combining information of two physical dimensions of spectrum and motion to identify a region of interest which is different from a background melt; extracting multi-dimensional features such as rheology, relative motion and spectral stability representing rigidity, motion independence and material stability of the region for the region; and carrying out collaborative analysis based on the characteristics to judge whether the artifacts are impurities or dynamic fluid artifacts. Through multi-dimensional feature collaborative analysis and physical judgment logic simulation, the problem that targets are difficult to distinguish only by means of single visual features is solved, and the accuracy and reliability of machine vision in intelligent impurity removal analysis application of non-ferrous metals are remarkably improved.