45results about How to "Reduce the impact of distracting factors" patented technology

The invention discloses a sea surface target detection method based on improved RBD significance calculation. The method comprises the following steps that (1) an original colored sea surface image isobtained; (2) a simple linear iteration clustering (SLIC) algorithm is used to carry out super-pixel segmentation on the sea surface image; (3) a semantic segmentation algorithm based on a Gaussian mixture model (GMM) is established, the image is segmented into sky and seawater areas, and a corresponding sea-sky line obtained by fitting boundary points of the sky and seawater areas into a straight line; (5) all super pixels are divided into seawater and sky types according to position information of the sea-sky line; (5) the significance is calculated according to super pixel characteristic information of the sea water type via the improved RBD (robustness background detection) algorithm, and a salient map of the sea water area is obtained; and (6) the salient map is segmented according to a fixed threshold, and a foreground target, namely the sea surface target, is extracted. Thus, the sea surface target can be detected from the complex background effectively, and the accuracy is higher.

The invention is suitable for the technical field of nuclear power station flyer monitoring, and provides an image processing method and a device, and terminal equipment, and the method comprises thesteps: detecting a region of interest in an image through employing a preset target detection algorithm; calculating the area of each region of interest, and screening the regions of interest of whichthe area is smaller than a preset area threshold as target regions; and performing target identification and / or target tracking on a target object corresponding to the target area. According to the image processing method, the problems of large calculation amount and low accuracy of image processing when a small flying object in a video picture is detected by using an existing image processing scheme can be solved.

The invention relates to a skin cancer type recognition method and system, computer equipment and a storage medium. The method comprises training a U-Net segmentation model and a DenseNet classification model; wherein the U-Net segmentation model forms a focus area mask; obtaining a focus area image;and identifying and classifying the lesion area images according to the extracted lesion classification feature information through a DenseNet classification model. The skin cancer symptom types are distinguished according to image recognition, the influence of other interference factors in the skin image is reduced, the extracted classification feature information is finer and more concentrated on a lesion part, and the skin cancer classification accuracy is greatly improved.

The invention relates to the wireless sensor network positioning field, and specifically relates to a wireless sensor network positioning method based on an RSSI (received signal strength indicator). The method comprises the following steps: a network is established; a host computer software sets a position coordinate of each reference device and a configuration parameter of each mobile device; the reference device sends a built-in RSSI value and an ID to other reference devices; when the reference device A serves as a difference correction reference node, a reference positioning coordinate and a positioning offset amount of the reference device A are calculated; the mobile device, according to received data, obtains a preliminary positioning coordinate and a difference correction coordinate of the mobile device; a correction mean value coordinate is obtained after repeated r rounds of above steps; and finally a weighted algorithm is used to calculate a positioning coordinate. The method of the invention solves the problems in the prior art of low positioning precision, low reliability and high cost, substantially reduces effects of the environment interference factor, and is suitable for the development trend of the indoor wireless positioning.

The invention provides a biosensor. The biosensor is used for detection of biological substances. The biosensor comprises 1, an electrode system having a complex stereochemical structure, 2, a reagent composition for corresponding biological substance detection and 3, a combined switch system for automatical identification of correcting codes. After being automatically sucked into the electrode system, a detected sample undergoes a biochemical reaction and an electrochemical reaction with the reagent composition to produce electrons; the electrons are transmitted to a tester by an electrode; the tester reads data and the data is processed; and a detection result is displayed. The combined switch system for automatical identification of correcting codes solves the problem of inconvenience of manual inputting of relevant information, avoids detection result errors produced by input errors, and has high production efficiency and a low cost.

The embodiment of the invention provides an information transmission and data processing method, device and system and a storage medium. In the embodiment of the invention, for the target informationdisplayed on a screen of the second terminal, when the first terminal needs the target information, the first terminal is matched with the second terminal; a screen of the second terminal is capturedby the second terminal; a screenshot operation of a second terminal is taken as a basis, information capable of indicating the area position of the target information on the screen of the second terminal and cross-screen transmission are combined, so that the first terminal can finally obtain the target screenshot containing the target information; compared with the mode of directly shooting thescreen of the second terminal, the influence of interference factors such as screen refresh frequency on the screenshot is relatively small, the quality of the screenshot is relatively high, and the definition of the information obtained by the first terminal is relatively high.

The invention discloses a registering method used for correcting wavelength drift of satellite hyperspectral data. The method is mainly used for carrying out wavelength registering on the satellite hyperspectral data with wavelength drift. The method is based on simulation reference data consistent with the baseline characteristics of the satellite spectral data. The method comprises the steps that the reference peak position of a satellite spectrum and a simulation spectrum is positioned by a trigonometric function of specific spectral resolution; initial registering is carried out; and through accurate inversion of a wavelength correction factor, a wavelength registering model is combined for fine registering. According to the invention, the method can automatically, accurately, stably and quickly acquire the wavelength drift of the satellite hyperspectral data without manual intervention or constructing a lookup table; the wavelength error of satellite observation is corrected; and the application value of the hyperspectral data in the fields of high precision atmospheric gas component detection and the like is improved.

The invention relates to a solar cell panel rainbow pattern detection method based on dual-light integration, and the method comprises the steps: obtaining an RGB image and a thermal imaging image ofa photovoltaic power station through a dual-light integrated camera, processing the RGB image of the photovoltaic power station, and obtaining a solar cell panel surrounding frame; according to the three-channel pixel value mean value corresponding to each solar cell panel surrounding frame and the thermal imaging image of the photovoltaic power station, rejecting the solar cell panel surroundingframes which do not meet the requirements, and carrying out temperature anomaly analysis on the solar cell panel RGB image surrounded by the solar cell panel surrounding frames which meet the requirements to obtain an abnormal RGB area; judging whether the abnormal RGB area is a strip-shaped area or not, and if yes, judging whether rainbow lines exist or not according to the pixel difference degree. The method combines the advantages of the RGB image and the thermal imaging image, takes the advantages of the RGB image and the thermal imaging image, greatly reduces the impact from an interference factor, and improves the recognition accuracy of the rainbow pattern of the solar cell panel.

The invention relates to the field of indoor wireless positioning and particularly relates to an RSSI (Received Signal Strength Indicator) wireless positioning method based on classified thresholds and signal strength weight. The method comprises the steps of: in a scene in an offline training stage, firstly collecting signal features of a plurality of reference points, establishing an RSSI fingerprint database, and then independently calculating a matching threshold for each reference point according to signal transmission loss features of the scene; and in a positioning stage, firstly preferably selecting the reference points from the fingerprint database according to fingerprints of a to-be-measured position and the threshold of each reference point, and then introducing signal strength of the reference points as the weight to participate in K adjacent weighting, so as to realize the positioning of a mobile node. The method is simple and feasible, is high in positioning accuracy and is capable of effectively reducing the influences of the random jitter of an RSSI and an individual abnormal beacon node on the positioning and the calculation.

The invention relates to a hydraulic impactor. The hydraulic impactor comprises a body, a drag device, a pressure isolation barrel, an impact body and an anvil, wherein the body is a hollow cylinder, the drag device is fixed at the upper part of the body, the anvil is fixed at the lower part of the body and extends into the body, the pressure isolation barrel is arranged between the anvil and the drag device, and the lower end of the pressure isolation barrel is meshed with the upper end of the anvil; the impact body is communicated with an inner cavity of the anvil, a nozzle is fixed in the anvil, the impact body is arranged in the pressure isolation barrel, the outer wall of the impact body is matched with the inner wall of the pressure isolation barrel, the lower end surface of the impact body is matched with the upper end surface of the anvil, and the impact body is restrained to perform reciprocating motion; a downward bulge is arranged in the center of the drag device and adopts a cylinder and cone combined body, the impact body is a barrel-shaped body provided with a step outside, and the bulge of the drag device corresponds to an upper end port of the impact body. Hydraulic energy is used sufficiently, a drill bit is enabled to generate axial vibration during rotation work, and the rock breaking efficiency is increased.

The invention provides a coal cutter hydraulic support side guard state monitoring method and device, and the method comprises the steps: carrying out the defogging of a continuous monitoring image of a coal cutter hydraulic support side guard, and carrying out the edge detail information recovery of the defogged monitoring image, and obtaining a fusion image; identifying a face guard plate area in the fused image through a preset target detection model to obtain a target image; the face guard in the target image is recognized to obtain the angle of the face guard, and the state of the face guard is determined according to the angle, the state of the face guard near the coal mining machine can be monitored in real time by preprocessing the monitoring image and recognizing the image, and the monitoring effect and reliability are improved.

The invention relates to a game testing method and device, electronic equipment and a computer readable medium, and belongs to the technical field of game testing. The method comprises the steps: obtaining an original scheme of a to-be-tested function of a game, and obtaining a to-be-tested scheme of the to-be-tested function according to the original scheme; determining a test server with a to-be-tested function, dividing users entering the test server into comparison users or test users in a test time period, and enabling the comparison users to play a game according to the original scheme,wherein the test users play the game according to the to-be-tested scheme; obtaining comparison data of the comparison user in the test time period and test data of the test user in the test time period, and obtaining a test result of the to-be-tested function according to the comparison data and the test data. According to the invention, the users in the test server are divided into the comparison users and the test users, and the corresponding original schemes or to-be-tested schemes are executed respectively, so the influence of interference factors in the test process can be reduced, and the test result is more accurate.

The invention discloses an air leakage position detection device for an engine cylinder. The detection device comprises a smoke generation device and a smoke detection device, and is characterized inthat the smoke generation device comprises a smoke box and a pressure measurement meter, an outlet of the smoke box is sequentially connected with the pressure measurement meter and the cylinder, andsmoke can be input into the cylinder; the smoke detection device comprises an endoscope, an LED lamp and a buzzer alarm, and the smoke detection device can extend into the air cylinder and is used fordetecting the smoke condition in the air cylinder and the aging or cracking condition of all parts; the detection device designed by the invention can realize rapid detection of in-cylinder leakage.

The invention discloses a multicellular sphere recognition and classification method based on deep learning. The method comprises the following steps: firstly, converting a multicellular sphere culture microscopic image into a grayscale image; training an RCNN algorithm model through a manually labeled training set to obtain an optimal RCNN detection model; recognizing and positioning target multicellular spheres in the image by the trained optimal detection model, and extracting sub-images of the multicellular spheres in the image; segmenting the cell spheres by adopting an algorithm based ona histogram segmentation threshold, extracting segmented multi-cell images, and training the segmented multi-cell images through artificially classified large-sample multi-cell sub-images to obtain an optimal cell classification model; and carrying out cell classification on the cell microscopic image by using the trained optimal cell classification model. According to the method, the cell recognition detection and generalization performance and robustness can be remarkably enhanced, the influence of an external environment and manual operation on a detection result is greatly reduced, the accuracy of detection analysis is improved, and meanwhile, the detection efficiency is improved.

The invention provides an impregnation automatic control system in production of short fiber stock solution viscose. The conventional automatic control system is improved and optimized, and a redundant process control system is constructed by redundant process control software, so that the impregnation automatic control system in the production of the short fiber stock solution viscose is established. The concentration of pulp is controlled by controlling the feeding speed and addition amount of the pulp and the addition amount of impregnation alkali; a feeding system comprises a pulp feeder and an impregnation alkali feeding pipe which are provided with devices for controlling the feeding amount of the pulp and the impregnation alkali; and the concentration of the pulp has been adjusted during feeding, so that the influence of other interference factors in the subsequent impregnation process is greatly reduced.

The invention discloses an aircraft cabin door jump prediction method and device, equipment and a medium. The method comprises the steps of obtaining a first feature point corresponding to a skeleton and a second feature point corresponding to a cabin door; acquiring a first shape error corresponding to the framework and a second shape error corresponding to the cabin door; according to the second feature point, obtaining a rotation speed vector and a normal vector corresponding to the second feature point; according to the rotation speed vector and the normal vector, obtaining an order difference change rate at the second feature point; and according to the first shape error, the second shape error and the order difference change rate, obtaining cabin door order difference prediction data. The method is based on the actually measured data of the aircraft cabin door and the skeleton, predicts the jump data in combination with the motion relation, more truly reflects the cabin door jump condition, does not need to assemble the cabin door on the skeleton for measurement, avoids repeated assembly, measurement and adjustment of the cabin door, is high in accuracy, and effectively improves the cabin door jump data measurement efficiency.

The invention relates to the technical field of chemical analysis, in particular to a method for measuring critical micelle concentration of a surface activity substance by a fluorescent probe method.The method uses n-B18H22 as a fluorescent probe substance, and it is found out that the fluorescence of the n-B18H22 can be improved by an anionic surfactant SDS. Under an optimal condition, fluorescent intensity and SDS concentration are respectively segmented within ranges of (0-8.2)*10<-3>mol / L and 8.2*10<-3>mol / L to 3.27*10<-2>mol / L, the surfactant concentration corresponding to an inflectionpoint of a curve is the critical micelle concentration (CMC), and therefore, a simple fluorescent method for measuring the CMC is built. A result shows that the CMC measured by the method conforms toa reported value, and the practicability and the accuracy of the method are declared.

The invention discloses a triggering device and an active system used for automobile rear-end collision safety protection measures. The triggering device used for the automobile rear-end collision safety protection measures comprises a housing, a triggering block, a spring and a contact switch, wherein the triggering block, the spring and the contact switch are positioned in the cavity of the housing; through mechanical motion response of the triggering device, the contact switch of the triggering device can accurately send a triggering signal in time when an automobile suffers rear-end collision at an automobile accelerated speed higher than a preset rear-end collision accelerated speed threshold ac; an active rear-end collision safety protection device is triggered to be started to provide protection measures for automobile passengers during the rear-end collision so that the passengers are protected to reduce damages; and moreover, the triggering device used for the automobile rear-end collision safety protection measures works on the basis of a mechanical motion response mode and is slightly affected by various interference factors; and the material and device cost of the triggering device is low and is only less than 1 / 10 of that of a high-accuracy collision sensor in the prior art; and therefore, the triggering device used for the automobile rear-end collision safety protection measures has the advantages of high stability, high reliability and low cost.

The invention relates to a solar cell panel RGB image identification method and device based on dual-light integration, and the method comprises the steps: obtaining an RGB image and a thermal imagingimage of a photovoltaic power station through a dual-light integration camera, processing the RGB image of the photovoltaic power station, and obtaining a solar cell panel surrounding frame; classifying the solar cell panel bounding boxes according to the mean value of the three-channel pixel values of each solar cell panel bounding box and a preset pixel threshold value, and combining the thermal imaging image of the photovoltaic power station and the target credibility of the related solar cell panel bounding boxes; and judging whether each solar cell panel surrounding frame surrounds the solar cell panel RGB image or not. According to the method, the advantages of the RGB image and the thermal imaging image are combined, the RGB image and the thermal imaging image are taken, the RGB image of the solar cell panel can be accurately and reliably recognized, the influence of interference factors is greatly reduced, and the recognition accuracy of the image of the solar cell panel is improved.

The invention discloses a method for rapidly detecting retrovirus by a cell ultrathin section electron microscope. The method comprises the following steps: 1) collecting and fixing of a sample; 2) uranium acetate block dyeing; 3) dehydrating, permeating and polymerizing; 4) slicing, dyeing and detecting through a transmission electron microscope. Compared with a traditional method, the method hasthe advantages that the sample preparation efficiency is greatly improved, the traditional 5-7-day sample preparation period is shortened to 1 day, and an ultrathin slicing technology for rapidly detecting viruses becomes possible. Compared with a traditional virus detection method, the optimized ultrathin section electron microscope technology for detecting viruses has the advantage that the specificity, accuracy and detection efficiency of virus detection are greatly improved.

The invention relates to the technical field of chemical analysis, in particular to a method for measuring critical micelle concentration of a surface activity substance by a fluorescent probe method.The method uses n-B18H22 as a fluorescent probe substance, and it is found out that the fluorescence of the n-B18H22 can be improved by an anionic surfactant SDS. Under an optimal condition, fluorescent intensity and SDS concentration are respectively segmented within ranges of (0-8.2)*10<-3>mol / L and 8.2*10<-3>mol / L to 3.27*10<-2>mol / L, the surfactant concentration corresponding to an inflectionpoint of a curve is the critical micelle concentration (CMC), and therefore, a simple fluorescent method for measuring the CMC is built. A result shows that the CMC measured by the method conforms toa reported value, and the practicability and the accuracy of the method are declared.

The invention discloses an edge rolling spinning pneumatic boosting system for coin making. The edge rolling spinning pneumatic boosting system comprises a machine body, a center disc and a check ring are arranged on the machine body, the center disc is connected with a corresponding rotation driving device, a plain edge gauge is arranged on the periphery of the center disc, a molding channel for spinning machining of a coin blank cake is formed between the center disc and the plain edge gauge, a cake conveying rotary disc is arranged at the bottom of the check ring, and a coin storage groove is jointly formed by the check ring and the cake conveying rotary disc. The cake conveying rotary disc is connected with a corresponding rotation driving device, a cake conveying outlet is formed in the lateral face of the check ring and communicated with the molding channel through a cake conveying sliding way, pneumatic spraying nozzles are formed at the cake conveying outlet of the check ring and the interior of the cake conveying sliding way and communicated with air sources through corresponding pipes, and the air spraying direction of the pneumatic spraying nozzles is coincident with the movement direction of the coin blank cake. By means of the edge rolling spinning pneumatic boosting system, two-section type dual boosting is carried out at the position of the cake conveying outlet and the position of the cake conveying sliding way, so that the edge rolling efficiency is improved by more than two times, and the edge rolling speed is stable.

The invention relates to an airborne camera exposure parameter adjustment method and system, and relates to the technical field of unmanned aerial vehicles. The method includes: obtaining the current environmental illuminance value by a photometer; according to the environmental illuminance value, determining the environmental exposure value of the airborne camera by using the environmental correspondence between the environmental illuminance value and the environmental exposure value of the airborne camera; according to the environmental exposure value, Determine the internal exposure value of the airborne camera by using the preset corresponding relationship between the environmental exposure value and the internal exposure value of the airborne camera; determine the internal exposure value of the airborne camera by using the corresponding relationship between the internal exposure value and the exposure parameters of the airborne camera The value of the exposure parameter; adjust the current exposure parameter of the airborne camera according to the value of the exposure parameter. The invention measures the environmental illuminance in real time according to the flying environment of the UAV, and adaptively determines the exposure parameters of the airborne camera through the correlation between the environmental illuminance and the exposure parameters of the airborne camera, thereby improving the accuracy of photos or spectral image information obtained by the airborne camera. accuracy.

The invention discloses a printed circuit board defect image registration method, device, equipment and storage medium. The method includes: obtaining a product picture and a preset template picture of a printed circuit board; Points are matched with the corresponding non-deformable key points in the preset template image to determine several non-deformable registration positions; several deformable key points in the product image are matched with the corresponding deformable key points in the preset template image, to determine a number of deformable registration areas; perform image transformation on the product picture according to the number of non-deformable registration positions and the number of deformable registration areas, and obtain a defect registration image of the printed circuit board. The invention reduces the influence of interference factors and improves the accuracy of defect registration images by separately matching non-deformable key points and deformable key points.

The invention discloses a video technology based real time stockyard airborne dust detection and intelligent watering dust depositing device and method. The device comprises a video monitoring device,a computer, a wind speed and wind direction measuring device, a PLC and a watering device. An optimized neural network or other algorithm is combined with static and dynamic characteristics of airborne dust to process real-time video data of stockyard airborne dust, an initial coordinate of a dust area and an estimated value of the airborne dust concentration are calculated, wind speed and wind direction measurement data is combined to implement calibration, the coordinate of the dust area, namely the coordinate of a watering area, is determined finally, and according to the coordinate of thewatering area and the estimated value of the airborne dust concentration, the PLC controls the watering device to carry out watering dust depositing in a dynamic graded way according to areas Thus, real-time monitoring can be realized in different environmental conditions, a control instruction is timely, and the construction and maintenance cost is low; and dust is inhibited effectively, it is ensured that increase of the water content of materials stored in the stockyard is within a controllable range, and labor, material and transport materials are saved.

The present application provides an image matching method based on edge direction and gradient features, which includes: training the offline template image, specifically: determining the high and low threshold parameters, denoising the template image; determining the number of template image pyramid layers and each layer Rotate the angle step; expand the template image and change the image pyramid; expand the image of each layer in the template image pyramid; extract the 0-angle feature from the expanded image; rotate the features of each layer image in the pyramid; use the trained offline template image to online The image to be tested is matched, specifically: image expansion is performed on the image to be tested; image pyramid changes are performed on the expanded image to be tested; rough matching is performed on the topmost image in the pyramid; fine matching is performed on the image to be tested after rough matching; The target points of the image to be tested are sorted and screened to obtain the coordinates, angles and matching scores of the target to be matched in the image to be tested. The application can perform image matching quickly and accurately.

The invention discloses a parameter determination method and a parameter determination device, and relates to the technical field of medical data processing. The parameter determination method comprises the steps that a first parameter set corresponding to a to-be-tested body is determined based on a first time interval, the first time interval comprises a night interval, and the first parameter set comprises a first heart rate characterization parameter and a first heart rate variability characterization parameter; based on a second time interval, a second parameter set corresponding to the to-be-tested body is determined, the second time interval comprises a daytime interval, and the second parameter set comprises a second heart rate characterization parameter and a second heart rate variability characterization parameter. The influence of interference factors in the parameter measurement process is reduced, and the purpose of improving the accuracy of the obtained parameter set and the comparability along with time change is achieved. In addition, when the first parameter set and the second parameter set determined by the invention are used for representing the level and / or the state of the neurohormone, the representation accuracy and the comparability along with time change can be effectively improved.

The invention relates to a dimension stone shape classification method based on a rapid PCA algorithm and a K-means clustering algorithm, and the method comprises the following steps: (1), carrying out the statistics of coordinates and gravity center coordinates of all edge contour points of dimension stones according to an earth-rock mixture section CAD graph which is obtained through a digital image processing technology and only has contour information; (2) unifying the position, the size and the orientation of the dimension stones; (4) adopting a rapid PCA algorithm to reduce 360 radiusesof each dimension stone in a dimension stone contour radius statistical table obtained by the unification processing to two dimensions by taking the 360 radiuses as row vectors; (5) selecting a K-means clustering algorithm to cluster the matrix after dimension reduction, so that dimension stones with similar shape features are classified into one class; and (6) finely adjusting the result after partial classification according to the defined shape characteristic parameters, and classifying the dimension stones with generality into one class. The method can quickly and accurately perform combined operation such as movement, rotation, scaling and direction changing on a large number of dimension stones, and efficient classification of a large number of dimension stones with different shapesis achieved.

The invention relates to a microbial electrochemical analysis device and an analysis method thereof, and belongs to the technical field of microbial electrochemical analysis. The invention discloses a microbial electrochemical analysis device, which is characterized in that an air cathode is arranged on one side of an electrolytic tank cavity, meanwhile, the device also comprises a counter electrode and a reference electrode, the device can be freely switched between a microbial electrode three-electrode analysis mode and a two-electrode maintenance mode, and a microbial electrode signal is accurately measured through the three-electrode mode; the long-term maintenance requirement of the microbial electrode is met through a two-electrode mode, and the defect that a traditional microbial battery structure cannot meet two requirements at the same time is overcome. The invention also discloses a microbial electrochemical test analysis method, which is characterized in that the microbial electrochemical analysis device disclosed by the invention is connected with an electrochemical workstation, so that the steady-state polarization curve of the microbial electrode and the resistance distribution analysis of the microbial electrode are realized; the problems of accurate characterization and efficient analysis of the performance of the microbial electrode are solved.