42results about How to "Improve efficiency and precision" patented technology

A positioning device is used to place an implant in the proper vertical position within the soft tissue and bone, provide space for subsequent procedures, and create a proper emergence profile for the final restoration. The lower portion of the positioning device has substantially the same outer shape as that of the implant but without external screw threads or other mechanical features for anchoring the implant in the jawbone. After the hole is countersunk, the depth of the hole in the jawbone is increased incrementally and checked with the positioning device for collar engagement and location of the margin within the gum tissue and bone. After the proper vertical position is attained, a special countersink bur may be used to create needed space around the implant margin and provide a proper crown emergence profile.

The invention discloses a novel point cloud parallel Softassign registering algorithm. On the basis that CUDA performs parallel acceleration on a Softassign algorithm, by use of a method of combining three-dimensional point cloud discrete curvature estimation and three-dimensional Kd-tree, point-cloud simplification is performed on a three-dimensional object point cloud to enable the simplified three-dimensional object point cloud to maintain sufficient geometric characteristics, and then Softassign registering is performed on the simplified object point cloud, such that the registering precision of the Softassign algorithm in three-dimensional object point cloud registering is improved. The algorithm provided by the invention has the following advantages: first of all, the point-cloud simplification is performed on the three-dimensional object point cloud, such that the information content of the Softassign registering is reduced; and secondly, through the improved Softassign registering algorithm, the registering precision in object cloud registering is improved, and through a parallel acceleration technology, the operation speed of the Softassign registering method is enhanced.

An embodiment of the invention discloses an image similarity confirmation method. The image similarity confirmation method comprises the following steps of confirming the first image confidence coefficient which is corresponding to at least one category; extracting at least one first visual characteristic of a first image; extracting at least one second visual characteristic of a second image; confirming the similarity of the first image and the second image at least according to the first image confidence coefficient which is corresponding to at least one category, the at least one first visual characteristic and the at least one second visual characteristic. The invention also discloses an image similarity confirmation device. The image similarity confirmation method and device can avoid the problem that the characteristic extraction in the prior art is lack of pertinence and improves the image similarity confirmation efficiency and accuracy.

The invention provides an operation system for puncture, a control method and a control device. The operation system for puncture comprises an imaging device, a mechanical arm and the control device,wherein the imaging device is used for acquiring a focus image of a focus organ, and a puncture part is arranged at the far end of the mechanical arm; the system also comprises a guide part, and the control device controls the puncture part to move along a preset puncture track through the mechanical arm and adjusts the preset puncture track according to state information of the guide part. The preset puncture track is adjusted by use of the state information of the guide part, the condition that a puncture operation robot cannot accurately puncture the focus points because of displacement ofa patient or one part of the patient is avoided, and the accuracy and the efficiency of puncture are improved.

The invention relates to a CT image liver tumor segmentation method based on deep learning, and specifically solves the problems that the original U-Net depth is small, the structure is simple, a model with a good training effect is difficult to train, and the accuracy of CT image liver tumor segmentation is insufficient. Specifically, the network structure of the U-Net is optimized and improved;a Dropout layer and a Batch Normalization layer are added to modify a network structure, a VGG-16 network module is used to replace an encoder part of U-Net, and an optimized model is established to perform accurate segmentation on a liver tumor CT image.

The invention relates to the field of mounting of optical devices and discloses a lens laminating system and method. The system comprises a precise location system; the precise location system comprises a spatial electric sliding table and an industrial camera component, the industrial camera component is arranged at the bottom of the spatial electric sliding table, the industrial camera componentis used for judging the exact location of a material box or base arranged below, a grabbing device is also arranged at the bottom of the spatial electric sliding table, and the material box is used for placing a lens. The lens laminating system and method which are provided by the invention have the advantages that the spatial electric sliding table is arranged, so that the grabbing device can bedriven to move in space, and the lens laminating system can automatically move to the material box and the base; the industrial camera component is arranged, so that locations of the material box andthe base can be precisely judged, and the grabbing device can effectively grab a lens and laminate the lens on the base; and the lens laminating system disclosed by the invention can automatically realize laminating of the lens, so that labor intensity is reduced and precision and efficiency are improved.

The invention discloses an integrating and conveying mechanism with the tracing function. The integrating and conveying mechanism with the tracing function comprises a tension servo motor (1) and a thrust servo motor (2) and is characterized in that the tension servo motor (1) drives a tension chain (5) fixedly provided with a tension integration conveying belt (4) by driving a tension chain wheel (3); the thrust servo motor (2) drives a thrust chain (8) fixedly provided with a thrust integration conveying belt (7) by driving a thrust chain wheel (6); the tension chain (5) and the thrust chain (8) are arranged in parallel so that the tension chain (5) and the thrust chain (8) cannot interfere with each other in the running process; and the tension integration conveying belt (4) and the thrust integration conveying belt (7) are arranged in a spaced manner. Under the joint action of the tension servo motor and the thrust servo motor, the tension integration conveying belt and the thrust integration conveying belt which are adjacent to each other mutually make tracing movement, the integrating and conveying efficiency and precision of materials are improved, and the integrating and conveying mechanism with the tracing function is suitable for being used and popularized.

The present specification provides a method, apparatus, server, and data processing method for determining a timeout period. The method for determining the timeout length comprises the following steps: obtaining the tolerant failure rate of the target system and the probability distribution of the call time consumption, wherein, the probability distribution of the call time consumption is obtainedby data fitting of the call time consumption length of the server call thread in the target system in a preset time period; determining a target time elapsed according to the tolerance failure rate and the probability distribution of the call time elapsed; a timeout time of a server in that target system being determined accord to the time consuming time of the target. In the embodiment of the present specification, the invocation time consuming probability distribution is obtained by fitting the invocation time consuming length of the server invocation thread in the preset time period in advance, and the timeout length is determined according to the probability distribution and the tolerant failure rate, so as to make thread invocation, thereby improving the accuracy and efficiency of determining the timeout length.

The invention relates to a power head structure of a four-corner corner connecting machine. The power head structure of the four-corner corner connecting machine comprises a corner connecting head base plate, a corner connecting head table plate, a concave positioning and side positioning integrating mechanism, a convex positioning mechanism and a perpendicular compression mechanism, wherein equally high supporting plates are perpendicularly installed on the corner connecting head base plate, the corner connecting head table plate is arranged on the equally high supporting plates, the concave positioning and side positioning integrating mechanism is horizontally installed on the corner connecting head table plate, the perpendicular compression mechanism is installed above the concave positioning and side positioning integrating mechanism, and the convex positioning mechanism is perpendicularly installed on the corner connecting head table plate. By effectively integrating original concave positioning, side positioning, convex positioning and perpendicular positioning, the power head structure of the four-corner corner connecting machine improves the efficiency of positioning and adjusting four-corner doors and windows, and improves the yield rate.

The invention discloses a semantic query method based on a medicine body library. According to the method, bodies in the medicine body library are constructed under the assistance of a field expert in reference to a skeleton method. The bodies are constructed through the following steps: capturing the bodies, namely, acquiring concepts and terminologies and relations thereof in the medicine field under the assistance of the field expert, and determining that the concepts and terminologies and the relations are unambiguous; structurally representing medicine; performing body evaluation: judging whether or not crude medicine bodies are consistent with a standard, if so, saving the bodies in the forms of files, otherwise returning to a body capturing stage; and finally, saving medicine body information in the forms of files.

The invention provides a liquid density monitoring device and method. The device comprises a floater part immersed in a liquid, a connecting component, a potentiometer and a computing component; a floater can float up and down along with variation of density of the liquid, the liquid is contained in a stable groove and has a preset height; the connecting component is located above the floater andis connected with the floater, and the connecting component can synchronously move with the floater; the potentiometer comprises a fixed end on the stable groove and an electrical contact piece contacted and conducted with the fixed end, the electrical contact piece is fixedly connected with the connecting component, the electrical contact piece can slide relative to the fixed end in the verticaldirection, and magnitude of current flowing through the poteniometer is changed along with variation of a relative position between the electrical contact piece and the fixed end; and the computing component is respectively connected with the fixed end and the electrical contact piece and jointly forms a circuit loop with the potentiometer, and the computing component is used for measuring a current value in the circuit loop and converting the current value into a density value. The liquid density monitoring device provided by the invention can monitor the density of the liquid in real time, and liquid density monitoring efficiency can be improved.

The invention discloses a content providing method, a terminal, a server and a system. The content providing method includes that the server receives relevant information of a currently-played first content transmitted by the terminal and acquires a second content and relevant information thereof according to the relevant information of the first content, and the second content and the relevant information thereof are transmitted to the terminal, so that the terminal displays the second content in relation to the first content on time according to the relevant information of the second content, and the relevant information of the second content is used for determining trigger conditions for displaying the second content. By the content providing method, the terminal is capable of automatically displaying the relevant content according to the content currently played, and accordingly, efficiency and accuracy in display of relevant contents are improved.

The invention discloses a use method of a face recognition device. The use method comprises the following steps: uploading face information of a user to a face library for storage; then the camera recognizing the face data, processing the face data by the detection module and the comparison module and then uploading the data to the central processing unit, and the central processing unit uploadingthe processed information to the face library to compare the data and judge whether the user is a registered user or not. The device has a multi-stage positioning correction function, can automatically select an adjustment mode according to the difference between the position corresponding to the high-definition camera and the face position, can improve the adjustment efficiency and precision, and can effectively improve the face positioning efficiency and precision.

The invention provides a fault detection system. The fault detection system comprises a tooling plate, a general control device and a first connector; the second connector in the tooling plate is connected with equipment to be detected; the signal generator in the general control device is connected with the first connector; when a test needs to be carried out, the first connector is controlled tobe in abutting joint with the second connector; and a signal processing module can notify a signal generation unit to send an audio and video signal to the equipment to be detected to test. By meansof the fault detection system provided by the invention, each worker does not need to connect and separate a line; furthermore, line connection can be carried out before the equipment to be detected enters production line flow; the production line operation is not influenced by the line connection process; artificial participation is unnecessary in the test process; therefore, the production lineoperation stability is improved; and thus, the detection efficiency and precision are greatly improved.

The invention provides a navigation broadcasting method and device, equipment and a storage medium. The method comprises the following steps: acquiring a current position of a navigation object; when determining that the navigation object is about to arrive at a preset maneuvering point according to the current position, acquiring environmental data near the maneuvering point; determining a navigation statement corresponding to the maneuvering point; determining broadcast parameters of the navigation statement according to the current position and/or the environment data; and performing voice broadcasting on the navigation statement according to the broadcasting parameters. According to the navigation broadcasting method provided by the embodiment of the invention, the broadcasting parameter of the navigation statement is flexibly adjusted based on the current position of the navigation object and the environmental data near the maneuvering point, so that the problems of redundant broadcasting, untimely broadcasting and the like can be avoided, the accuracy and efficiency of navigation broadcasting are improved, and the user experience is improved.

The invention discloses a bar finished product warehouse-out label identification method based on machine vision. The method comprises the steps of carrying out the target detection of an image of aninterested region through a pre-trained bundled bar detection model, recognizing a plurality of bundled bars, and outputting the position information of each bundled bar; based on the position information of the plurality of bundled bars, identifying labels on the plurality of bundled bars through a pre-trained label detection model, and outputting the position information of each label; and basedon the position information of the label, identifying the label to obtain label information. According to the invention, real-time detection can be carried out on a continuously running automobile automatic delivery production line, the material specification information is identified and automatically transmitted to other systems, manual detection is replaced, and the material specification information identification and check efficiency and accuracy are improved.

The invention provides the field of the mechanical equipment and provides a fastener loading mobile device which comprises a rack and a work table fixed on the top of the rack and further comprises amobile part arranged on the top end of the working table and a loading plate fixed on the top end of the mobile part. According to the fastener loading mobile device, a fastener is disposed in a loading opening, and a fasten baffle is inserted in the loading opening to fix and limit the fastener initially, then a fasten rod is used for secondly fixing and limiting the fastener to guarantee the fasten cannot move in the process, so that a first servo motor can control the loading plate to move, and process efficiency and accuracy is improved, the improvement of processing efficiency is facilitated.

The embodiment of the invention discloses a text classification method, a device, equipment and a computer readable storage medium. The method comprises the steps that firstly, a target corpus set of a target text is obtained, the target text comprises questions, abstracts and keywords, and the target corpus set comprises abstract corpora and question and keyword corpora; the target corpus set is imported into the trained text representation model to obtain a first feature vector of the target text, the first feature vector is used for reflecting a subject, an abstract and a keyword of the target text, then the first feature vector is imported into a first classifier to obtain a classification result of the target text, the classification result comprises a first classification category and confidence, and the first classification category is used for classifying the target text; and if the confidence coefficient is greater than or equal to a first threshold value, determining that the first classification category is the category of the target text. According to the embodiment of the invention, the accuracy and efficiency of text classification can be improved.

The invention discloses a supporting block detection device for a printing machine. The supporting block detection device comprises a fixed support, a two-dimensional code scanning gun, an infrared sensor, a label plate with a two-dimensional code, a printing supporting block and a detection controller which are arranged in the printing machine. The two-dimensional code scanning gun and the infrared sensor are arranged on the fixed support, the printing supporting block is arranged opposite to the fixing support, the label plate is arranged on the printing supporting block, a two-dimensional code with printing supporting block information is arranged on the label plate, and the detection controller is connected with the two-dimensional code scanning gun and the infrared sensor. By means ofthe mode, the supporting block detection device for the printing machine can detect and judge whether the current printing supporting block meets the machining requirement or not in real time, the positions of the code scanning gun and the sensor can be adjusted in a multi-dimensional mode so as to be matched with printing supporting blocks of different specifications and sizes, and the accuracyand high efficiency of scanning and detection are improved.

The invention discloses a steel plate defect detection method and a computer readable storage medium. The method comprises the following steps of: acquiring target measurement data of a steel plate; performing Gaussian filtering on the target measurement data to obtain filtering data; performing gradient calibration on the filtering data to obtain calibration data; performing seam calculation and levelness adjustment on the calibration data to obtain levelness data; and performing defect detection and deformation calculation on the levelness data according to preset plate shape defect conditions to obtain defect deformation parameters of the steel plate. According to the invention, the technical problem that steel plate defect online detection cannot be effectively realized in the prior art can be solved.

The invention discloses an unmanned aerial vehicle tree diameter measuring method based on an optical flow method. A tree board, an unmanned aerial vehicle and data processing equipment are used as measuring devices to measure and calculate the tree diameter, and the tree board is used for marking the number of a tree and the measuring position of the diameter at breast height; the unmanned aerial vehicle is provided with a monocular lens and an ultrasonic sensor, the monocular lens is used for shooting tree boards and tree trunks, and the ultrasonic sensor is used for keeping the unmanned aerial vehicle and the tree boards at corresponding heights; the data processing equipment is used for planning the route of the unmanned aerial vehicle, detecting tree boards in videos shot by the unmanned aerial vehicle, identifying numbers of the tree boards, and measuring the diameter at breast height of a tree trunk where the tree boards are located based on an optical flow method to complete tree diameter measurement. According to the method, the tree diameter can be measured quickly and accurately in batches in a high-automation mode, the method is easy to operate, manpower and time are greatly saved, the efficiency and accuracy of large-scale tree diameter measurement are improved, a complete tree diameter database is established, management and calling are convenient, and according to the tree diameter measuring and calculating method based on the optical flow method, the calculation precision is guaranteed while only a small number of calculation resources are needed.

The invention discloses an image display method and device, a storage medium and equipment, and the method comprises the steps: obtaining an acceptance image based on an image selection operation when the image selection operation is detected in a visual acceptance toolbar, and carrying out the acceptance image on an application image displayed on a terminal interface, and performing overlapped display on the acceptance image in a default transparency mode, obtaining an image adjustment operation for the acceptance image, and controlling the acceptance image to perform image adjustment display based on the image adjustment operation. By adopting the method and the device, the acceptance image is displayed on the application interface of the terminal in an overlapping manner, so that the purpose of visual acceptance is achieved, the image is more real and visual, and the efficiency and the accuracy of visual acceptance are improved.

The invention provides a statistical method and device of litigation case information, equipment and a storage medium. According to the method, a case statistical request is sent to a server in response to a case summarization operation on a graphical user interface, the case statistical request comprises at least one query time period and a user identifier, and a statistical result returned by the server is received, wherein the statistical result is obtained by analyzing the processing condition according to the time limit information of all cases corresponding to the user identifier, and the statistical result is displayed on the graphical user interface. The cases are counted through the electronic platform, the statistical efficiency and accuracy are improved, and the statistical result at least comprising one statistical index is processed to ensure the high efficiency of statistical information.

The invention relates to the technical field of video editing, and discloses a sports event video editing method and system based on artificial intelligence. An original video is segmented and marked through a generated event time table, a user logs in a segmented video and label storage server through an intelligent tactical tablet for playback, and the video is edited from a complete game. According to the method, each attack video paragraph is rapidly extracted, and corresponding video tags are generated, so that searching and watching after the competition are facilitated, and a user can conveniently, rapidly and specifically analyze each key time of a complete competition, thereby realizing research and analysis on effective paragraphs and tactics of competition teams. The efficiency and the accuracy of event video editing are greatly improved.

The invention relates to the technical field of beach garbage treatment, and particularly discloses a beach garbage treatment method based on intelligent machine equipment and data processing, and the method comprises the following steps: S1, dispatching a beach garbage cleaning robot to enter a working area; s2, garbage in the working area is collected; s3, the collected garbage is conveyed to a garbage transfer recycle bin; s4, battery charging is conducted on the beach garbage cleaning robot through the machine energy supply station; and S5, the beach garbage cleaning robot with the replaced battery enters the working area again to work. The invention provides a new beach garbage treatment scheme based on intelligent equipment and cloud intelligent computing for the first time, provides a solution for beach environment deterioration, and establishes an integrated service system comprising a garbage transfer recycle bin, a machine energy supply station, a control center and a cloud data processing center. The monitoring and control of beach facilities are realized, and the guarantee is provided for the work of the intelligent robot.

The invention discloses a deep learning-based diabetic nephropathy early prediction method and system. The method comprises the following steps of: S1, collecting a subject sample, and extracting proteomics and lipidomics; S2, analyzing the proteomics and the lipidomics to obtain primitive characters; S3, screening the primitive characters, and extracting fusion features; S4, analyzing correlation among the fusion features to obtain a biomarker for early prediction of diabetic nephropathy; and S5, based on the biomarker for early prediction of diabetic nephropathy, conducting early prediction of diabetic nephropathy. According to the deep learning-based diabetic nephropathy early prediction method, massive and high-dimensional data volume calculation is carried out by using deep learning, sensitive correlation characteristics between diabetes and nephropathy are found, and the early prediction efficiency is improved.

The invention discloses a soft soil microstructure sample preparation device which comprises a sampling device, a liquid nitrogen heat insulation tank and a peristaltic pump, wherein the sampling device and the peristaltic pump are respectively communicated with the liquid nitrogen heat insulation tank through hoses. When in operation, the sampling device is inserted into soil by closing and opening a locking mechanism, a microstructure sampling bin device reaches a designated position, a valve of the peristaltic pump is opened, liquid nitrogen in the liquid nitrogen heat insulation tank is pressed into a liquid nitrogen cavity by pressure, thermo sensitive color-changing volatile liquid in a closed temperature sensing cavity changes along with soil temperature, color change of the outer wall of the cavity is displayed by observing outer wall temperature sensing materials of an upper cavity, a soil sample is taken out by the microstructure sampling bin device, and sampling is finished. Operation of manual sample cutting is avoided, sampling efficiency and accuracy are improved, cost is effectively saved, experimental efficiency is improved, the soil sample is not destroyed, and microstructures of different positions of the soil are observed by determining the position of the cut soil sample.