48results about How to "Detection speed is slow" patented technology

The invention discloses a dynamic malicious software detection method based on a virtual machine and sensitive Native application programming interface (API) calling perception. The method consists of three parts, namely a Xen secondary development-based analysis and detection environment, a monitoring control program, and a training learning and detection program of a malicious software classifier. A detection model is divided into a training stage and a detection stage, wherein the training stage comprises the following steps of: executing a sample set file in a clean analysis environment for fixed length time, acquiring Native API calling frequencies of process behavior, privilege behavior, memory behavior, registry behavior, file behavior and network behavior of the sample set file, and training the classifier by using the data; and the detection stage comprises the following steps of: executing files to be checked, counting the Native API calling frequencies of six sensitive behaviors in the fixed length time, and classifying the detected files by using the trained classifier to obtain classification results which are malicious software or normal files. The method is still effective for the malicious software with anti-virtual, anti-debugging and anti-tracking capabilities.

An image display apparatus with image entry function of high accuracy enabling high-speed direct screen input without decreasing a pixel aperture ratio. Data lines of thin-film transistors which do not receive light and storage lines are connected to respective selector switches. The selector switches are turned on and off by a switching signal supplied through a switching line from a control circuit. The conveyance of a drive signal and a video signal supplied from a gate line driving circuit and a data line driving circuit and the conveyance of a light signal to an X address detection circuit and a Y address detection circuit are switched by turning on and off the selector switches.

A specific clothing image identification and detection method based on machine learning is provided. The invention relates to an image identification and detection method. The purpose of the invention is to solve the problems of the prior art that a manual detection way cannot realize searching of images of specific kinds of clothing because of a large quantity. The method includes the following processes: firstly, identifying image contents, and obtaining identified image contents; constructing an image classification database, processing images in the image classification database through a self-service sampling method, and obtaining processed images; obtaining an optimized convolution neural network; carrying out model integration; obtaining N individual learning devices, and combining the N learning devices through a simple ballot way; and secondly, utilizing a Faster R-CNN method to detect the identified image contents obtained in the first step. The specific clothing image identification and detection method is applied to the clothing image identification and detection field.

A laser machining apparatus that includes an interferometric height position detection unit, a condensing point position adjustment unit, and a confocal optical height position detection unit. A controller of the laser machining apparatus specifies, as a correction value, the difference between the two height positions of the upper face of the workpiece, one detected by the confocal optical height position detection unit and the other detected by the interferometric height position detection unit. The controller controls the condensing point position adjustment unit based on the height position obtained by correcting the height position of the upper face of the workpiece detected by the interferometric height position detection unit using the correction value.

The invention belongs to the technical field of computers, and discloses a lightweight improved target detection method and system, and the method comprises the steps: replacing an original backbone network in an FSSD with an improved ShuffleNet v2 model; introducing a weighted bidirectional feature pyramid structure to replace an original three-layer feature fusion structure in an FSSD, carryingout top-down and bottom-up feature fusion on multiple layers of features, so that the fused features contain more semantic information, and target detection is performed. Compared with a basic model,the model provided by the invention has the advantages that the accuracy is improved, the detection speed is reduced, and the weight of the model is increased to a certain extent. Through the analysisand the performance of experimental data, the improvement of the accuracy rate can be seen to accord with the improvement provided by the invention, and meanwhile, the defects in the aspects of detection speed and model weight caused by the improvement of the invention are also within a predictable range.

The invention discloses a method for testing the water content of foamed asphalt, and relates to material testing or analysis with nuclear magnetic resonance. The water content of the foamed asphalt is measured by a nuclear magnetic resonance device. The method comprises the following steps: preparing a foamed asphalt test sample of which the water content is to be tested; setting a nuclear magnetic parameter of the nuclear magnetic resonance device; acquiring a nuclear magnetic resonance semaphore of each gram of asphalt from which impurities are removed; determining a relationship between the water content of a standard test sample and a nuclear magnetic resonance semaphore; acquiring a total nuclear magnetic resonance semaphore of the foamed asphalt test sample of which the water content is to be tested; acquiring the mass of foamed asphalt evaporation residues in the foamed asphalt test sample of which the water content is to be tested; calculating the water content of the foamed asphalt test sample of which the water content is to be tested. According to the method, the defect of incapability of a method in the prior art to measurement of the water content of the foamed asphalt and the defects of an industrial CT method, disclosed in CN103558236A, for testing the water distribution of an asphalt mixture are overcome.

The invention relates to a target detection method in a mining truck environment based on improved tiny-yolov3. The method comprises the following steps: S1, acquiring image data of a target object; S2, preprocessing the acquired image data of the target object; S3, inputting the preprocessed image data of the target object into the tiny-yolov3 model, processing through the tiny-yogov3 model to obtain pixel position coordinates of the target object in the image; wherein the tiny-yolov3 model is a model improved by combining a residual network structure, and the target object image data is thetarget object image data obtained under the overlook condition. According to the detection method provided by the invention, the detection precision of the target can be greatly improved on the premise that the operation speed is not reduced.

The invention discloses a precision detection device and method for multi-core cable terminal crimping quality. The device comprises a detection speed control circuit; the detection speed control circuit is connected with a detection timing sequence control circuit; the detection timing sequence control circuit is connected with an open circuit and dislocation detection circuit; the open circuit and dislocation detection circuit is connected with a failure judgment and warning circuit; the detection timing sequence control circuit is also connected with a sampling, comparison and amplifying circuit; the sampling, comparison and amplifying circuit is connected with an A / D conversion circuit; the A / D conversion circuit is connected with both a storage / counting control circuit and a CPU; the storage / counting control circuit is connected with both the CPU and a liquid crystal display module; the CPU is connected with the liquid crystal display module. The precision detection device and method for multi-core cable terminal crimping quality has the advantage that failures, such as open circuit, dislocation and excessive crimping resistance, occurring after terminal crimping of multi-core cables having more than eight cores can be detected and judged precisely and automatically.

The invention discloses a live pig ingestion behavior analysis method. The method comprises the following steps: acquiring group cultured live pig videos, and constructing a data set; carrying out preprocessing and data enhancement on the target data set, manually markingthe position frame and category information of the ingested pigs, and carrying outK-means clustering anchorfor the pig form; andtraining by using an improved YOLOv3-tiny algorithm, wherein the improved YOLOv3-ty model is a model obtained after dense connection improvement is carried out in combination with a ResNet thought and a small target detection scale is added; acquiring a to-be-tested live pig ingestion image and inputting the ingestion image into the trained model for detection to obtain a result image. Compared with a method before improvement, the detection precision is improved on the basis of ensuring certain detection real-time performance.

An aflatoxin detection reagent kit based on a loop-mediated isothermal gene amplification method is based on the basic principle of loop-mediated isothermal amplification (LAMP). The loop-mediated isothermal gene amplification method includes particular steps: preparing a culture medium of PDA (potato dextrose agar); cultivating strains; designing a loop-mediated isothermal gene amplification (LAMP) primer; extracting a fungus DNA (deoxyribose nucleic acid) template; and proportionally preparing, adding a PCR (polymerase chain reaction) tube and realizing water bath heat-insulation reaction, agarose gel electrophoresis and the like. The aflatoxin detection reagent kit is combined with a method (GB\T13092-2006) for detecting the total number of mold in feed, the concentration of selected fungus liquid is detected by an MPN (most probable number) method, and the aflatoxin detection reagent kit is prepared according to the method. When used for detecting aflatoxin fungi in the feed and food, the reagent kit is high in specificity and sensitivity, quick in detection and low in cost, can be used for detecting a plurality of samples at the same time, and is applicable to sanitation examination for primary-level feed and worthy of being widely popularized.

The invention relates to the technical field of semiconductor manufacturing, and specifically discloses a positioning aligner. The positioning aligner comprises a rotating shaft and four claws. The claws are arranged on the rotating shaft, the claws are successively arranged to encircle the rotation center of the rotating shaft, and the angles formed between each two claws are respectively 45 degrees, 135 degrees, 45 degrees and 135 degrees; a through space is formed in the direction of the 135-degree included angles for transmitting a wafer by a manipulator; and the radius of a circle formed by encircling the external peripheries of claws is smaller than the radius of the wafer. According to the invention, the aligner for detecting eccentricity of a bonding pad by use of a claw structure, compared to a conventional claw type aligner, a carriage structure is cancelled, the rotating shaft does not have to ascend and descend, and an eccentricity detection function can be completed. At the same time, since an ascending and descending process is cancelled, the detection speed is greatly improved, and through properly reducing the rotation speed, wafer slip can be prevented.

The invention relates to a surface defect detecting device based on laser diffuse scattering. The device comprises a laser, a beam expander, a first adjusting mirror, a second adjusting mirror, a third adjusting mirror, an octagonal prism, a telecentric scanning lens, an inner integrating sphere, a measured surface and a first photoelectric sensor; the first adjusting mirror, the second adjustingmirror and the third adjusting mirror are sequentially disposed behind the beam expander, the octagonal prism is disposed above the third adjusting mirror, the center of the octagonal prism is provided with a driving shaft, a first opening is formed in the bottom of the inner integrating sphere, a second opening is formed in the upper portion, and the inside of the integrating sphere is provided with the first photoelectric sensor. The surface defect detecting device based on laser diffuse scattering adopts laser scanning and a scattering detecting technology, can detect a smooth surface in ashort time and determine the specific location of a defect, can perform preliminary processing on a defect image, achieves visual operation, and has the advantages of fast scanning speed, high detection efficiency, high precision and the like.

The invention discloses a method for managing threads, and is applied to product detection. The method comprises the following steps: step S1, monitoring whether an idle thread or a new product detection task exists at present; step S2, if the idle thread exists but the new product detection task does not exist at present, enabling the idle thread to be dormant; step S3, if the idle thread does not exist but the new production detection task exists at present, judging whether the dormant thread exists at present; step S4, if the dormant thread exists at present, awaking the dormant thread at present, and distributing the awaken thread to the new product detection task; and step S5, executing the step S1 until product detection is finished. The invention further discloses a device for managing the threads. The resource occupancy rate of detected products can be reduced, and hen the product detection speed is increased.

The invention discloses an intelligent elevator door state detection method based on deep learning. Through the analysis and recognition of a video monitoring image in an elevator car, the elevator door state is detected. Firstly, a target detection neural network model improved based on a yolov3-tiny framework is used for conducting target detection on an elevator door according to a certain period, and the location information of an elevator door target frame is detected, updated and saved. Every time when judging the elevator door state, aiming at a current elevator image, a picture is cutaccording to the saved location of the elevator target frame, then a user-defined efficient classification network model based on deep learning is used for classifying and analyzing the state of an elevator target frame picture, according to the three states of 'open', 'closed' and 'not closed well', a confidence probability value is given, and the state classification is conducted.

The invention discloses equipment and a method for detecting defects of a boiler water cooling wall tube on the basis of a low-frequency electromagnetic technique. The equipment comprises a power supply module, a processor, a touch screen, an excitor, a scanning probe, a signal preprocessing circuit and a power supply module, wherein the signal preprocessing circuit is used for preprocessing a signal received by the scanning probe; the power supply module is used for supplying power; the output end of the processor is connected with the input end of the scanning probe through the excitor; the output end of the scanning probe is connected with the input end of the signal preprocessing circuit; the processor is connected with the signal preprocessing circuit and the touch screen. By adopting the equipment and the method, detection on internal defects of the boiler water cooling wall tube can be rapidly and accurately completed, and the equipment and the method are applicable to sealed tubes.

The invention provides a device and a method for measuring the content of norfloxacin-based on terahertz light. The device comprises a terahertz instrument, a terahertz light spectrum emission terminal, a terahertz light spectrum reception terminal and a treatment terminal, wherein the terahertz light spectrum emission terminal is used for making a terahertz light spectrum generated by the terahertz instrument act on a sample; the terahertz light spectrum reception terminal is used for receiving the terahertz transmitted wave information of the sample; the treatment terminal is used for calculating the content of the norfloxacin in the sample according to the frequency characteristic of the norfloxacin absorption characteristic of a terahertz waveband in the obtained terahertz transmitted wave information. Through collecting the terahertz light spectrum of the sample, the characteristic curve of the light spectrum of a norfloxacin-containing sample is obtained; according to the characteristic curve, the frequency characteristic of the norfloxacin absorption characteristic of the terahertz waveband is extracted; these properties are combined; the quantitative analysis modeling is carried out; thus, the content of the norfloxacin in the sample in obtained; the measurement is accurate; the sampling amount of the sample is small; the detection speed is quick; the economic loss is decreased.

The invention relates to a non-uniform resampling technology-based maneuvering target coherent integration detection method. According to the method, demodulation and pulse compression processing areperformed on received radar receiving signals in a fast time domain; fast Fourier transformation is performed on pulse compressed signals; constraint conditions are calculated, and an empty matrix isgenerated; whether a selected time point and selected time delay meet the constraint conditions at the same time is judged, if the selected time point and selected time delay meet the constraint conditions, non-uniform resampling is performed on range-direction frequency-domain signals, corresponding element values are calculated, if the selected time and selected time delay do not meet the constraint conditions, the corresponding element values are 0; all time points and time delay are sequentially selected, generated element values are arranged into the matrix; fast Fourier transformation and inverse fast Fourier transformation are performed on correlation function matrix data, so that coherent integration is completed; and whether a target exists or not is judged by means of a coherentintegration matrix. The method disclosed by the invention is not influenced by a searching method in a detection process, is low in calculation complexity and excellent in detection performance.

The invention discloses an elevator car personnel intelligent recognition method based on deep learning. By identifying and analyzing the images in the elevator car, the number of people taking the elevator in the elevator car and the state that whether people taking the elevator wear masks or not are detected. Whether someone is in the lift car or not is quickly determined through a simple and efficient image classification convolutional neural network. For a person state, a target detection neural network is used to detect each person in the elevator car and simultaneously determine "mask wearing", "non-mask wearing" and "head" persons. If the person who does not wear the mask is found, reminding voice is played through the camera in the lift car. Meanwhile, the number of people taking the elevator in the elevator car is counted for flow statistics and people crowding or overload reminding.

The invention discloses a water quality mutation early warning and monitoring system and method for a drinking water source site. The system comprises a spectrum detecting probe, an optical fiber, a spectrometer, a central processing unit and a display, the spectrum detecting probe is used for spectrum testing of a water body to obtain a spectrum signal, the spectrum signal is transmitted from theoptical fiber to the spectrometer for spectrum recording, the spectrometer is connected to the central processing unit through a data line, a spectrum recorded by the spectrometer in the central processing unit is compared with a standard spectrum in a database, the concentration of pollutants in the water body is analyzed, the concentration analyzed of the pollutants is used for judging whethera certain pollutant exceeds a standard, and the concentration analysis result in the central processing unit is displayed through the display. The system can achieve rapid detection without secondarypollution, and can achieve multiple parameters, especially simultaneous monitoring of unknown target pollution.

The invention relates to the technical field of mobile communication equipment, and provides mobile phone detection equipment for detecting strength of a mobile phone screen by utilizing a power pressing rod. The equipment comprises mobile phone detection equipment, driving frames are movably connected to the left side and the right side of the mobile phone detection equipment, and extrusion blocks are movably connected to the left side and the right side of the mobile phone detection equipment through the driving frames. And the left side and right side of the mobile phone detection equipmentare movably connected with adjusting frames through the driving frames According to the mobile phone detection equipment for detecting the strength of the mobile phone screen by utilizing the power pressing rod, the mobile phone screen is manually placed in the center of the outer surface of the power pressing rod by a worker, a switch of the equipment is turned on, and electric energy drives thepower pressing rod to move towards the inner side, the power pressing rod drives stretching frames on the left side and the right side to move towards the two sides, torque springs limit the moving positions of the stretching frames, the stretching frames drive the movable receiving rods to move towards the inner side while moving, and then the movable receiving rods drive the switches outside the connecting frame rods to be turned on.

The invention relates to a method for detecting the pearlite content of ductile cast iron. N representative ductile cast iron samples (200-300*phi24) are prepared and are heat-treated by using different heat treatment processes to obtain the ductile cast iron with different matrix structures; the pearlite content of each sample is detected by using the traditional metallographic detection method, and then is detected by an eddy current detection method. The detection results show that the pearlite content of the ductile cast iron and eddy current values have a linear relationship, and the error values of the two detection methods are within 3%, and can meet the accuracy requirements of industrial application.

The invention provides an operation control and island detection method and system for a photovoltaic direct-current converter. The method comprises the steps of 1, enabling the direct-current converter to achieve the maximum power tracking control of photovoltaic power generation in a mode of double-closed-loop control of a voltage outer loop and a current inner loop; 2, superposing an external small-disturbance sinusoidal signal on the voltage reference value or the current reference value; 3, sampling a DC bus voltage signal; 4, enabling the DC bus voltage signal to pass through a band-passfilter and a feedback gain to form a feedback disturbance quantity; 5, superposing the feedback disturbance quantity on the external small-disturbance sinusoidal signal, and superposing the feedbackdisturbance quantity with the original voltage or current reference value to form a new voltage or current reference value; 6, applying the new voltage or current reference value to voltage or currentcontrol; and 7, detecting the DC bus voltage, and when the disturbance deviates from the preset operation range, judging that the DC power distribution network is in an island state. According to theinvention, the electric energy quality and the operation efficiency are improved.

The invention provides a low-altitude flyer detection method based on a compressed SSD, and relates to the technical field of target detection. The channel pruning algorithm is adopted to compress the model, and the detection speed of the model is increased. The method comprises the following steps: firstly, adding a batch normalization layer in a basic model, and introducing a scale factor for each channel; then, a scale factor is jointly trained, so that the numerical value of the scale factor is associated with the importance of the channel. And finally, judging the importance of the channels by measuring the size of the scale factor, and cutting off the channels with low importance. The performance of each aspect of the compressed model is superior to that of the original model, the network structure of the model is more refined, the size is greatly reduced, the operation speed is higher, and a higher detection speed can be obtained. According to the channel pruning method adopted by the invention, the importance of each channel is measured, so that the influence of the pruned channels on the precision of the model is small, and therefore, the precision of the optimized model is close to that of the original model.

The invention relates to a depth counterfeit image detection method fusing depth and width learning, which comprises the following steps of: firstly, converting an image to be detected into a gray level image, segmenting the gray level image into two groups of image blocks, extracting a centralized frequency domain magnitude spectrum of each image block, applying an attention mechanism to the centralized frequency domain magnitude spectrum, and carrying out channel connection on the two groups of image blocks to obtain a primary feature; secondly, constructing a channel convolution self-encoding module, performing feature fusion on the primary features by using an encoder of the pre-trained channel convolution self-encoding module to obtain two intermediate features, and respectively taking the two intermediate features as inputs of a feature mapping stream and a feature enhancement stream to obtain two mapping features and an enhancement feature; and finally, constructing three classifiers according to a width learning system principle, and carrying out weighted average on output results of the three classifiers to obtain a final detection result. According to the method, the attention mechanism is applied to the image block, the region with obvious tampering traces can be concerned from global information, data and time required by model training are less, and both accuracy and efficiency are realized.

The invention relates to an electrochemiluminescence enzyme-linked immunosensor and application thereof, the electrochemiluminescence enzyme-linked immunosensor is prepared by the following method: firstly, synthesizing mesoporous silica on an ITO electrode, then attaching the mesoporous silica electrode to a paper cover plate, then allowing a biotin labeled antibody and an alkaline phosphatase labeled antibody to react with an antigen to form an antigen-antibody complex; adding avidin magnetic beads to separate the antigen-antibody complex, finally adding a substrate disodium phenylphosphateto react, and detecting the generated phenol, so that the content of the to-be-detected substance is in direct proportion to the generated phenol, and theelectrochemiluminescence enzyme-linked immunosensor is constructed The electrochemical luminescence enzyme-linked immunosensor can be applied to detection of different clinical biomarkers with different concentrations, can simply, conveniently and rapidly adjust the detection range, has a lower detection limit and a wider detection range, and has a wide application prospect in the field of simultaneous detection of multiple components with large content differences.

The invention discloses an allocation processing method which is used for detection threads and applied to product detection. The allocation processing method comprises the steps that 1, a fixed number of detection threads for executing a product detection task are established; 2, whether a new product detection task exists or not at present is monitored; 3, if the new product detection task exists at present, whether the total number of the detection threads being executing the product detection task at present reaches the fixed number or not is judged, and a judging result is obtained; 4, allocation processing of the detection threads is conducted according to the judging result; 5, the step 2 is executed and quits until product detection stops. The invention further discloses an allocation processing device for the detection threads. According to the method and device, the resource occupation rate of detected products can be decreased, and then the product detection speed is increased.

The embodiment of the disclosure relates to a method and apparatus for detecting the state of a switch. The apparatus comprises a transformer, a trigger, and a controller. A primary winding of the transformer is coupled to the switch electrically; the trigger coupled to a secondary winding electrically is configured to respond to a trigger signal and generate a response signal associated with the switching state of the switch; and the controller is configured to generating the trigger signal and generating a switching signal for expressing the state of the switch based on the response signal.

The invention discloses a high-sensitivity specific troponin I latex turbidimetric detection kit adopting a transmission-scattering integrated method, and belongs to the technical field of biological detection. The invention discloses a high-sensitivity specific troponin I latex turbidimetric detection kit adopting a transmission-scattering integrated method. The kit comprises a first reagent R1 and a second reagent R2, the R1 is prepared from an HEPES-NaOH buffer solution, NaCl, polyethylene glycol 12000, Tween-20, a rabbit polyclonal antibody blocking agent and an oxidizing agent; r2 comprises a buffer solution, tween-20, carboxyl latex of 480 nm, carboxyl latex of 380 nm, a protective agent, a troponin I rabbit polyclonal antibody, a SignalUpTMWestern detection sensitizer and recombinant human serum albumin. The detection rate of healthy people on the surface of the kit is larger than or equal to 95%, the deviation of transmission and scattering results is smaller than 10%, the existing turbidimetric transmission defect is overcome, and the reagent detection rate is remarkably increased.

The invention relates to a pointer type pressure instrument reading method based on machine vision. The pointer type pressure instrument reading method comprises the following steps of obtaining an instrument standard template image and an instrument image to be read; matching the to-be-read instrument image with the instrument standard template image by using a matching algorithm to obtain a corrected to-be-read instrument image; denoising the corrected to-be-read instrument image, performing threshold segmentation on the denoised corrected to-be-read instrument image to obtain a background-removed to-be-read instrument image, and positioning the position of a boundary circle of a dial plate in the background-removed to-be-read instrument image; performing expansion operation on the background-removed instrument image to be read to obtain a pointer image; obtaining a needle point coordinate (x1, y1) and a circle center coordinate (x2, y2) in the pointer image; and obtaining the angular position of the pointer according to the tip coordinates (x1, y1) and the center coordinates (x2, y2), and obtaining the reading of the instrument according to the angular position of the pointer. Compared with the prior art, the method is advantaged in that reading of an instrument is more accurate.