58results about How to "Avoid culling" patented technology

The invention provides a non-texture spatial object attitude estimation algorithm based on contour point ORB (Oriented FAST and Rotated BRIEF) feature matching, which belongs to the technical field ofdigital image processing. On the basis of three-dimensional model prior, data information contained in the contour of a projected image is made full use of, ORB feature matching and color indexing are used to build a 2D-3D feature corresponding relationship from an input image to an object three-dimensional model, matching errors are used to build a confidence probability matrix, and a weighted orthogonal projection algorithm is put forward to calculate six-degree of freedom attitude parameters of the non-texture spatial object. The ORB features improve the accuracy of contour point matching,and certain robustness is realized in a condition of large offset of an initial attitude relative to a real attitude. The contact between building of the 2D-3D corresponding relationship subproblem and calculation of the attitude parameter subproblem is thoroughly explored, the matching errors are used to build the confidence probability matrix as the prior information for calculating the attitude parameters, the phenomenon that an RANSAC algorithm is used to eliminate error matching points is avoided, and the algorithm calculation efficiency and the precision are improved.

The invention discloses a binocular vision positioning method and system in a dynamic environment applied to the technical field of robot positioning. The method comprises the following steps of extracting image feature points, distinguishing the feature points on the moving object points and the feature points on the non-moving object points in a mode of clustering the feature points. The featurepoints on the moving object can be effectively eliminated so that elimination of feature points on a static dynamic attribute object is avoided, more available feature points are reserved, and poseswhich are closest to IMU data and are estimated based on different feature point sets are screened through an inter-frame pose transformation matrix obtained through an IMU pre-integration algorithm,so that some external points such as feature points belonging to a moving object are effectively eliminated. According to the poses and depth estimation method integrating monocular motion vision andbinocular stereoscopic vision, the estimation precision can be effectively improved.

The present invention relates to a road environment visual perception method based on an improved Faster R-CNN. Aiming at high-precision requirements of target detection and identification in a complex road scene, the invention provides an improved Faster R-CNN algorithm based on multi-GPU training. According to the algorithm, a multi-GPU parallel training method is used, so that the training efficiency is improved. By employing the ResNet-101 feature extraction network, the target detection precision is improved. By employing the Soft-NMS algorithm to reduce the omission factor, at the same time, introducing an OHEM into the ROI NetWork to reduce the false alarm rate, in order to improve the target detection effect of the algorithm in rainy, snowy and haze weather, the model is trained bycombining an internationally recognized automatic driving data set KITTI and Oxford RobotCar. The experimental results prove that by comparing with the Faster R-CNN, the algorithm has the advantagesthat the training speed and the detection precision are obviously improved, and particularly, and has good generalization ability and stronger practicability in an automatic driving scene.

A method for eliminating dense stellar field in a catalogue of stars comprises (1) selecting n stars with stellar magnitude lower than MT from an astronomical catalogue with gross star number of t to form a pre-selection catalogue of stars, (2) arranging a score counter with initial value of zero for each star in the pre-selection catalogue of stars, (3) selecting a neighborhood omega i of the i star, (4) detecting omega i, when j star exists in omega i and stellar magnitude Mj of the j star is lower than stellar magnitude Mi of the star i, then star i score counter Count i +1, (5), repeating the step (3) and step (4) until all stars in the pre-selection catalogue of stars pass through the same treatment, (6) traversing the pre-selection catalogue of stars, eliminating all stars with score counter larger than zero in the pre-selection catalogue of stars, and forming a navigational catalogue of stars with the rest stars. The method for eliminating dense stellar field in a catalogue of stars resolves the problem that two close stars are both selected as navigational stars, is favorable for increasing the uniformity of navigational stars distributed in the whole celestial sphere, and favorable for increasing star identification probability of star sensor.

The invention relates to a high-resolution remote sensing image inclined ship target detection method based on position sensing. The method comprises the following steps: extracting a multi-scale depth feature map by using a UNet-like multi-scale convolution network; directly predicting the possibility score that each anchor point has a ship and the position of the anchor point in a prediction boxthrough a rotating box regression model based on the anchor points by utilizing the extracted depth feature map; and correcting the anchor point score by using the position sensing score correction model. According to the method, the UNet-like convolutional neural network is utilized to extract both deep semantic features and shallow detail features, the model positioning precision is improved while the classification precision is ensured, and the detection performance of small-size ships is improved. And meanwhile, the candidate box score is further corrected by utilizing the position sensing score correction model, so that the candidate box positioning precision is improved.

The invention relates to the technical field of biomedicine, in particular to a method for preparing a humanized ACE2 mouse model. A mouse resistant to SARS-CoV2 can be infected with SARS-CoV2, so that the mouse can be used as a COVID19 infected disease model. Due to the advantages of short growth cycle and low cost of the mouse, the research and development speed of anti-coronavirus drugs and vaccines can be quickly promoted.

The invention discloses a data processing method based on a multilayer information united index. The method comprises the steps that a data transmission data index comprising virtual channel frame counts, virtual channel identifiers and frame time is extracted from original data; virtual channel data is extracted according to the data transmission data index, and a virtual channel data index comprising the virtual channel frame counts, the virtual channel identifiers and the frame time is established; source packet data is extracted according to the virtual channel data index, and a source packet data index comprising the virtual channel frame counts, source packet counts, source packet identifiers and source packet time is established; and sequencing, splicing and repetition-removing operation are performed on the source packet data according to the source packet data index to generate continuous and complete source packet data. Through the method, the problem that the data with identical virtual channel frame counts and source packet counts will be deemed as repeated data and removed through traditional methods is avoided, the continuity and completeness of the data are guaranteed, and the processing demand of space science on satellite sounding data can be met.

The invention discloses a shooting method and a shooting device. The shooting method comprises the following steps: detecting whether ambient light is suddenly changed in the shooting process, if so, pausing shooting or not using the currently acquired image; when the ambient light returns to normal, recovering the shooting or returning to use an subsequently acquired image, thereby avoiding acquiring the image with abnormal background or removing the image with abnormal background, the image with abnormal background can be prevented from synthesizing to the final synthesis image to influence the picture quality, the interference of the suddenly changed ambient light on the shooting process is eliminated, and the shooting effect is improved.

The invention discloses a magnetic disc fault handling method and device. The method comprises the steps of monitoring each magnetic disc in a magnetic disc array in real time, and acquiring fault information of a failure magnetic disc when a fault is detected to occur in an arbitrary magnetic disc; and acquiring a current state of the magnetic disc array, judging whether the arbitrary magnetic disc is allowed to be removed by the magnetic disc array according to the current state, if yes, removing the failure magnetic disk, otherwise continuing to use the failure magnetic disc and restoring the fault of the fault magnetic disc according to the fault information. When the failure magnetic disc is monitored, the failure magnetic disc is not directly removed, while the failure magnetic disc is removed when the current state of the magnetic disc array is determined to allow removal of a magnetic disc, thus, the magnetic disc is prevented from being directly removed when the current state of the magnetic disc array is not allowed to remove the magnetic disc, the risk of data loss caused by collapse of the whole magnetic disc array is greatly reduced, and the data reliability is ensured.

The invention relates to a box-packed chemical transmission, separation, coding and sorting system. The system comprises a transmission device, a separation device, a coding device and a sorting device which are arranged in sequence, wherein the transmission device is arranged on one side of the outlet of a packing device, and comprises a transmission cylinder and a conveyor belt; the transmission cylinder is mounted at the other side of the outlet of the packing device; the conveyor belt is mounted on conveyor rollers; partition plates are mounted on the conveyor belt at equal intervals; the separation device comprises a sinking separation conveyor belt; the coding device is arranged on the separation conveyor belt; a photoelectric detection switch is arranged on the coding device; the sorting device is arranged at the output end of the separation conveyor belt, and comprises a laser scanning light, an elimination cylinder and an elimination conveyor belt; the laser scanning light is arranged right above the separation conveyor belt; the elimination cylinder is arranged on one side of the separation conveyor belt; and the elimination conveyor belt is arranged on the other side of the separation conveyor belt. According to the system provided by the invention, the moving posture of a box-packed chemical, the identification of chemical specification, and precise laser coding can be guaranteed.

The invention provides a print quality detection method, a computer device and a computer readable storage medium. The print quality detection method comprises the following steps of acquiring the production information of a print, specifically, the target material information of the print is recorded in the production information; and acquiring the on-site material information of the print to beproduced, and judging whether the on-site material information is consistent with the target material information, and if not, sending out the warning information. The computer device and the computerreadable storage medium can achieve the print quality detection method. By means of the print quality detection method, the computer device and the computer readable storage medium, the appearance ofan incomplete product caused by the inconsistency between the print material and the preset target material can be avoided, and the quality of the print is monitored through a visual inspection modeafter each printing process is monitored, so that the detection efficiency of the print quality is improved.

The invention discloses a novel fruit thinning method. A near-infrared spectroscopy is utilized to scan sucrose solutions of different concentrations, characteristic wavelengths are extracted, a mathematical equation of sucrose concentration and near infrared spectrum reflection strength under the characteristic wavelengths is established, and thus a method for detecting content of sucrose in carpopodium of a fruit is obtained by adopting the near-infrared spectroscopy. The near-infrared spectroscopy is adopted to scan the carpopodium of the primarily-born fruit on the plant to obtain near-infrared spectrum data, and the concentration of the sucrose in the carpopodium is obtained according to the mathematical equation in the first step; the concentration of the sucrose in the carpopodium is compared with the concentration of sucrose in carpopodium of an adjacent fruit, and the fruit with the lowest concentration of the sucrose in the carpopodium is selected to thin fruits. According to the method, the fruit thinning effectiveness can be improved, and the method can be applied to fruit thinning of tomatoes, sweet peppers and eggplants and the like.

A bag arranging mechanism for a bag making system comprises a support assembly. The support assembly comprises an upper bottom plate and side plates on the left side and the right side. A certain number of polished shafts are arranged between the two side plates, and a first conveying assembly, a front bag arranging assembly, a rear bag arranging assembly and a second conveying assembly are installed between the side plates. According to the bag arranging mechanism for the bag making system, the conveying stability of packaging bags is guaranteed, the packaging bags are stacked, deviation andoverturning are avoided, the number and the width of conveying lines can be adjusted according to needs, adjustment is quite convenient, the mechanism is suitable for the packaging bags of various sizes, and unqualified inferior-quality packaging bags can also be effectively removed.

The invention relates to a file transmission method, file transmission equipment and a computer readable storage medium. The file transmission method is operated on file transmission equipment, and comprises the following steps: in response to a control instruction sent by terminal equipment, establishing control link connection with an FTP server; responding to a download file designation instruction sent by the terminal equipment, and establishing data link connection with the FTP server; sending a file downloading instruction to the FTP server to control the data link connection to start afile downloading operation; and in the file downloading process, controlling to send a control link connection keep-alive instruction to the FTP server at a preset time interval. Thus, in the file transmission process, the link connection keep-alive instruction is sent at the preset time interval until file downloading is completed, and the control link can be prevented from being removed by the server due to the fact that idle time is too long in the file downloading process.

The invention provides a power equipment online sensing and focusing algorithm based on deep learning, and the algorithm comprises the following steps: controlling an unmanned plane to carry out the fixed-point cruise, carrying out the online detection and analysis of a collected image, and adjusting the shooting angle and focal length of a camera according to an analysis result; according to thepower equipment online perception and focusing algorithm based on deep learning, the position of typical power equipment in an image is captured by using a deep learning means; whether the camera focal length is set properly is analyzed by using a focusing evaluation function, the shooting angle and focal length of the camera are adjusted by integrating the two information to obtain a clear imageof the equipment in the image center, the whole process of the algorithm is completed and adjusted online by the unmanned aerial vehicle, manual intervention is not needed, and the efficiency and quality of inspection work can be improved.

The invention relates to a control method for controlling an air source heat pump to be defrosted with a suction temperature. The temperature at the suction end of a heat pump system evaporator is detected and recorded; when the change rate of a detected value is larger than the preset value, a heat pump system begins to execute a defrosting program. By means of the control method, precise judgment on the real-time work condition of the heat pump system can be achieved just by detecting and judging the temperature at the suction end of the heat pump system, and thus the start time node of the defrosting program of the heat pump system is judged accurately. The invention furthermore relates to a heat pump water heater, and the heat pump water heater adopts the control method for controlling the air source heat pump to be defrosted with the suction temperature. Meanwhile, the heat pump water heater is simple in structure, the method is simple, the effect is obvious, and the water heater and the control method are suitable for being applied and popularized.

The invention discloses a bottle body output mechanism. The bottle body output mechanism comprises a guide device connected with an infusion bottle detection apparatus, a removing device for removingunqualified bottle bodies, and a conveying belt, the conveying belt is fixed to one side of the guide device, the guide device is used for conveying the bottle bodies towards the conveying belt, and the removing device is fixed between the guide device and the conveying device. The bottle body output mechanism can improve the production and processing efficiency and is further beneficial for relieving the labor intensity of workers and improving the product quality.

The invention relates to the technical field of data processing, in particular to a model data processing method which comprises the following steps: providing a main model X; p2, testing the main model X by utilizing the characteristic data to obtain a test result; p3, selecting feature data needing to be further expanded or constructed according to effect distribution feedback of thetest resultin the feature data value; p4, constructing the selected feature data needing to be further expanded or constructed to obtain new feature data; p5, retraining the main model X by using the newly constructed new feature data to obtain a model X1, and screening the constructed new feature data; p6, updating the feature data set by utilizing the screened new feature data to obtain an updated featuredata set; and returning to the step P2 again until the model Xn meets the preset stop condition. According to the effect distribution feedback of the test result in the feature data value, the featuredata needing to be further expanded or constructed is selected, so that the screened feature data has a better improvement space for the main model.

The invention discloses a conveying and arranging mechanism of a bag arranging machine. The conveying and arranging mechanism can be used for arranging and conveying packaging bags, and comprises twoarranging wall plates which are parallel to optical shafts and supported by the optical shafts, a transmission shaft a is erected at the front ends of the optical shafts, one end of the transmission shaft a is connected with a synchronous pulley a, and a driving wheel a connected with a motor is fixed to each arranging wall plates; and at least one arranging conveying assembly is arranged, the arranging conveying assembly comprises an aluminum tube beam, the aluminum tube beam is erected on the optical shafts a in a front-higher and rear-lower mode, the two sides of the front end and the rearend of the aluminum tube beam are provided with synchronous conveying wheels correspondingly, and the transmission shaft a passes through the two synchronous conveying wheels at the front end. An oscillating shaft is arranged, at least one pair of oscillating plates are connected with the oscillating shaft, an eccentric shaft and a shaft sleeve are arranged at one end of the oscillating shaft, andan oscillating motor is arranged on one sides of the arranging wall plates. According to the conveying and arranging mechanism, the number and height of the arranging conveying assembly body can be adjusted as needed, adjusting is very convenient, the structure is suitable for packaging bags with various sizes, the stability of conveying and arranging is guaranteed, and the packaging bags are guaranteed to be neat.

The present invention discloses an anesthesia depth and cerebral blood oxygen saturation monitoring device. A center place of an inside of a movable shaft is fixedly connected with a center rod, the center rod penetrates through elliptical blocks, an axis place of an inside of a support ring is fixedly connected with a center shaft, an outer ring of the center shaft is fixedly provided with clamping teeth, an inner ring place of the support ring is fixedly connected with an elastic inner ring, a bottom part of the clamping teeth is fixedly connected with a pipeline, and the pipeline is internally provided with abutting blocks and combining blocks in a penetration manner. Liquid absorbs temperatures and transmits the temperatures to the elastic inner ring, the elastic inner ring extrudes out movable plates from clamping holes, a passage inside the support ring is open, after cold water is rushed into the pipeline at bottom parts of the clamping teeth, the pipeline is softened, after theabutting blocks and combining blocks are affected by water flow to be rushed together to directly touch a button, an alarm connected outside the pipeline sounds to remind medical personnel to cool down and do further inspection for patients, and besides, after the cold water rushes into the pipeline, the cold water can help the patients to relieve heat and conduct physical cooling.

The invention provides a classification method and device based on a combined model, equipment and a storage medium, and can be applied to the classification fields of social contact, games, multimedia, vehicle connection, driving, shopping and the like. The method comprises the steps of obtaining m groups of training samples and n groups of prediction data; training the m models through the m groups of training samples to obtain m target models; obtaining a first classification label of each training sample in the first training sample group according to the target model, the first training sample group corresponding to the target model and the classification threshold of the first training sample group; obtaining a classification label variance corresponding to the target model accordingto the first classification label of each training sample and the second classification label included in each training sample; obtaining a combined model according to the m target models and the classification label variances corresponding to the m target models; and according to the n groups of prediction data, the combination model and the classification thresholds corresponding to the n groups of prediction data, obtaining a classification label of each prediction data so as to improve the accuracy of the classification method.

The embodiment of the invention provides a cache data migration method and device, and relates to the field of computers. The method includes the steps of when the free space of a calculation area is not enough to run a subtask dispatched by a Spark execution engine, and calculating the size of preemption space, wherein the size of the preemption space is the size of the running space of the subtask minus the size of the free space of the calculation area; calculating the sizes of the free space and effective space of a storage area, and judging whether or not the size of the preemption space is larger than the size of the free space of the storage area and the size of the effective space of the storage area; when the size of the preemption space is larger than the size of the free space of the storage area and is not larger than the size of the effective space of the storage area, migrating the cache data of the storage area to second level cache, wherein the size of the cache data migrated to the second level cache is the size of the preemption space minus the size of the free space of the storage area. The method and the device avoid the repeated operations of part of calculation tasks and protect the performance of the Spark execution engine.

The invention relates to the technical field of big data, particularly to an information analysis method based on big data. The method comprises: S1, publishing data by a user; S2, receiving big datainformation, and transmitting to cloud storage; S3, sending to a screening server; S4, screening data information; S5, sending the data to a server, and storing the data in resource disks in a classified manner; S6, submitting a resource acquisition request by the user; and S7, extracting information corresponding to the resource disk by the server, and submitting the information to the hand of the user. According to the information analysis method based on the big data, after the information analysis method is screened by the server for the first time, the opportunity of screening and auditing the documents again is achieved, five times of manual auditing are achieved, unqualified document cloud disks after screening can be saved for three months and then abandoned, use humanization is improved, manual auditing is conducted on disputed documents, the situation that compliant high-quality copywriting cannot be published is avoided, copywriting suspected to be copied repeatedly is eliminated, and it is avoided that many same copywriting appears in a manufactured webpage or server.

The invention relates to the technical field of agricultural machinery, in particular to a cotton delivering pipeline illuminating device. The cotton delivering pipeline illuminating device comprises a support part and a parabolic lampshade arranged on the support part, wherein the parabolic lampshade comprises a first parabolic segment, a straight line segment and a second parabolic segment which are connected in sequence. Light rays emitted by a first lamp tube located on an intersection point of the first parabolic segment and the second parabolic segment are respectively reflected by the first parabolic segment and the second parabolic segment to form parallel light parallel to symmetric axes of parabolas, the parallel light is irradiated into a cotton delivering pipeline from different directions, simultaneously light rays emitted by a second lamp tube located at the straight line segment is reflected by the straight line segment and then is irradiated into the cotton delivering pipeline, the light rays enhance illumination intensity and illumination uniformity of the cotton delivering pipeline, and shadow inside the cotton delivering pipeline is avoided. In addition, a distance between the parabolic lampshade and the cotton delivering pipeline can be adjusted through the support part, the illumination intensity of the cotton delivering pipeline can be further adjusted, and a good daylighting environment is provided for detection and removal of cotton foreign fiber.

The invention discloses a dyeing and finishing process of cold-feeling tencel linen, and belongs to the technical field of textile fabric production. According to the dyeing and finishing process, ina singeing process, dipping treatment can be innovatively performed on irregular hairiness on a fabric, and the hairiness is dried after absorbing a nano ferromagnetic solution, and endowed with magnetism; during singeing, a hairiness absorbing ball arranged at a fire hole is used for generating a magnetic attraction effect on the fabric, so that on one hand, the hairiness is forced to be highly separated from the fabric, the singeing effect is improved, the damage to the fabric is reduced, and on the other hand, the hairiness can be forced to be in full contact with inner flame, thereby avoiding underburnt influence caused by over-low temperature, and meanwhile, the damage to the fabric and energy consumption of gas brought by high temperature of outer flame are avoided. Compared with anexisting singeing process, the process has the advantages that the smoothness effect of the fabric can be remarkably improved, the pressure of subsequent procedures is reduced, a finished product is obtained through cold pad-batch, mercerization, dyeing, enzyme treatment, resin treatment, drying and tentering setting in sequence after singeing is finished, and the dyeing and finishing effect is greatly improved.

The invention discloses a grinding machine with a fixing function. The grinding machine structurally comprises a grinding machine body, a worktable, a baffle, a fixing mechanism, a driving device anda polishing mechanism, and the fixing mechanism is composed of an L-shaped support, a fixing clamping block, a clamping mechanism, a lead screw and a nut. The worktable can clamp workpieces of different shapes through the clamping mechanism, the clamping efficiency of the workpieces is improved, damage caused when the workpieces are fixed and clamped by the clamping mechanism can be avoided through a rubber mat, the situation of polishing offset can be avoided through the structural design, the overall polishing quality is improved, the safety of workers is improved, a grinding wheel spindle drives an air induction mechanism to pull surrounding waste debris to avoid the situation that the waste debris flies around, and consequently the adverse reaction is caused to human bodies by the waste debris existing in air, and then waste debris on the surface of the grinding wheel spindle is removed through a removing mechanism to prevent the situations that excessive impurities exist on the surface of the grinding wheel spindle, consequently accelerated deterioration of grinding oil, damage to parts and replacement of the grinding oil due to deterioration are caused, and the production cost is increased.

The invention, which belongs to the technical field of combination of surface enhanced Raman spectroscopy detection and chemometrics, discloses a method for predicting the content of each pesticide ina mixed pesticide solution based on surface enhanced Raman spectroscopy. The method comprises the following steps: respectively preparing a plurality of kinds of pesticides which do not chemically react with one another into solutions with uniform concentrations, then mixing the pesticide solutions according to different volume ratios, then mixing the mixed pesticide solution with a synthesized surface enhanced Raman substrate, and finally, collecting a surface enhanced Raman spectrum of the mixed pesticide solution; screening characteristic spectral band intervals of the collected surface enhanced Raman spectra of the mixed pesticide solution; and finally, predicting the content of each pesticide in the mixed pesticide solution by using the optimal surface enhanced Raman spectrum characteristic spectral band interval obtained by screening. The method has the advantages of high speed, high detection precision, good stability, small calculated amount and the like, and can be used for rapid and high-precision prediction of the content of each pesticide in the mixed pesticide.

The invention relates to a tobacco bale subpackaging device of a packaging machine. The tobacco bale subpackaging device comprises a subpackaging device controller, a front conveying toothed belt assembly, a rear conveying toothed belt assembly, a front and rear conveying toothed belt tensioning block, a motor and a driving device. Real-time subpackaging work can be achieved, the situation that a small box appearance detection system removes connected cigarette packets is avoided, the wrong removal rate of the small box appearance detection system is reduced, and material consumption is reduced. The distance between the front conveying toothed belt and the rear conveying toothed belt can be adjusted, the cigarette packet passing distance can be adjusted according to different cigarette packet marks, and the subpackaging function is achieved. The device is provided with the controller, different subpackaging speeds are set so as to be matched with different packaging machine speeds, and the subpackaging effect is guaranteed. The tobacco bales are conveyed through the tooth-shaped belts, the positions of the tobacco bales between the front tooth-shaped belt and the rear tooth-shaped belt are fixed, and quality defects caused by mutual extrusion of the tobacco bales in the subpackaging process are avoided.

The invention discloses a novel fruit thinning method. A near-infrared spectroscopy is utilized to scan sucrose solutions of different concentrations, characteristic wavelengths are extracted, a mathematical equation of sucrose concentration and near infrared spectrum reflection strength under the characteristic wavelengths is established, and thus a method for detecting content of sucrose in carpopodium of a fruit is obtained by adopting the near-infrared spectroscopy. The near-infrared spectroscopy is adopted to scan the carpopodium of the primarily-born fruit on the plant to obtain near-infrared spectrum data, and the concentration of the sucrose in the carpopodium is obtained according to the mathematical equation in the first step; the concentration of the sucrose in the carpopodium is compared with the concentration of sucrose in carpopodium of an adjacent fruit, and the fruit with the lowest concentration of the sucrose in the carpopodium is selected to thin fruits. According to the method, the fruit thinning effectiveness can be improved, and the method can be applied to fruit thinning of tomatoes, sweet peppers and eggplants and the like.