34results about How to "Meet the need for speed" patented technology

The invention discloses a privacy-protection index generation method for mass image retrieval, relates to the privacy protection problem in mass image retrieval and involves with taking privacy protection into image retrieval. The method is used for establishing an image index with privacy protection, and therefore, the safety of the privacy information of a user can be protected while the retrieval performance is guaranteed. The method comprises the steps of firstly, extracting and optimizing SIFT (Scale Invariant Feature Transform) and HSV (Hue, Saturation and Value) color histogram, performing feature dimension reduction by use of a use of a manifold dimension reduction method of locality preserving projections, and constructing a vocabulary tree by using the dimension-reduced feature data. The vocabulary tree is used for constructing an inverted index structure; the method is capable of reducing the number of features, increasing the speed of plaintext domain image retrieval and also optimizing the performance of image retrieval. The method is characterized in that privacy protection is added on the basis of a plaintext domain retrieval framework and the inverted index is double encrypted by use of binary random codes and random projections, and therefore, the image index with privacy protection is realized.

The invention relates to the field of violation behavior detection, and specifically discloses a construction violation behavior detection model construction method and application thereof. The methodcomprises the steps: carrying out the similarity pairing of images in an original image sample set, carrying out the foreground exchange of each group of image pairs, carrying out the style renderingof all exchanged images, and obtaining an extended image sample set; training a single-stage target detection network based on an original image sample set and an extended image sample set, adjustingand optimizing network parameters according to the text information of the images and prediction results of the category prediction branches and the position prediction branches in the detection network during training, and finally detecting violation behaviors based on the constructed violation behavior detection model. According to the method, pairing similarity enhancement is carried out on original samples, and the scale of training samples and the model training precision are greatly improved; in addition, context information is considered during training, regularization is carried out by adding the context information to the network, and the accuracy of target detection is greatly improved.

The invention discloses a robot and a control method. The robot comprises a body, a moving mechanism, an emergency stop mechanism and a body adjusting mechanism, the moving mechanism drives the body to move, the emergency stop mechanism comprises a connecting frame, a sliding rod, a sucker seat, a compression spring, a sucker, a cylindrical cam, a pulley shaft, a pulley, a driven stop gear, a driving stop gear, a stop motor and a solenoid valve, and the body adjusting mechanism comprises a fixed frame, a fixed column, a pitching motor seat, a pitching motor, a primary framework, a secondary framework, a rack, a telescopic motor and a rack driving gear. The emergency stop mechanism can achieve the emergency stop function of the robot under emergency circumstances such as pedestrians, animals and carts in the rapid movement process of the robot, and safety accidents caused by knock into the pedestrians and the like are prevented. The emergency stop mechanism can further prevent the robotfrom turnover due to over-high gravity center and dynamic change in the cargo handling process.

The invention discloses non-interrupted linear-speed packet transmitting and receiving methods and equipment based on LINUX, and belongs to the technical field of data packet transmitting and receiving in LINUX systems. The packet receiving method comprises the steps that an application program carries a serial number of at least one free buffer block to trigger an inner core module and a network card to perform data packet exchange; the inner core module exchanges the serial numbers of the free buffer blocks with a serial number of at least one stored data packet buffer block received in the network card, and configures physical addresses of the free buffer blocks to the network card; and the application program processes data according to the serial numbers returned by the inner core module. Therefore, the network card stores data packets into the buffer blocks which can be accessed by the application program during packet receiving; the data packets are transmitted by the physical addresses of the data packet buffer blocks filled by the application program during packet transmitting; data copy is not required during packet transmitting and receiving courses; and the transmitting and receiving performance of the data packets is improved. In addition, non-interrupted circulating active packet transmitting and receiving are adopted, so that calling of a CPU (Central Processing Unit) resource during packet transmitting and receiving interruption is avoided, and the packet transmitting and receiving performance is further improved.

The invention discloses a method and a system for realizing data caching in an offline state of equipment. The method comprises the following steps: monitoring a connection state of a communication link between the equipment and a background; when the connection state of the communication link between the device and the background is an abnormal connection state, caching data to be transmitted inthe device to a local first-level cache space for storage according to a preset two-level storage mode; And transferring the data stored in the first-level cache space to a local second-level storagespace according to a preset data transferring rule, and releasing the storage space occupied by the transferred data in the first cache space. In the two-stage storage mode, the first-stage cache space is a high-speed cache; The device has the advantages of being high in storage speed and small in storage capacity, the second-level storage space adopts a data base storage mode for low-speed storage, the second-level storage space has the advantages of being large in storage capacity and low in storage speed, data loss can be effectively prevented, and equipment and background communication data transmission is safer and more reliable.

The invention provides a Linux configuration based routing acceleration system and method of CPE, and belongs to the field of CPE. When the routing acceleration system is loaded for the first time, existing route information and interface information in a Linux protocol stack are read, the route information is written in an FIB of an adaptation layer, and the interface information is written in an LINK table of the adaptation layer; and information of a core is converted into item information of the FIB, and synchronized into the FIB and/or LINK table of the adaptation layer, the FIB and/or LINK table are/is maintained via a hook processing unit, a learning event unit, an equipment state unit and an update aging unit, and a route is forwarded according to the FIB and/or LINK table. The FIB (Forward Info Base) and the LINK table (output information table) are established and maintained in the adaptation layer, routing data is forwarded via the FIB and the LINK table, the load of the CPE is reduced, and the route forwarding speed is improved.

The invention discloses a TOA mode-based high-precision rapid positioning method in a non-line-of-sight environment. The method comprises the steps of obtaining a basic positioning equation between amobile station and each base station according to a distance observation value; selecting a corresponding basic positioning equation between the mobile station and the base station when the distance observation value is minimum as a reference equation, and subtracting the reference equation from basic positioning equations between other mobile stations and the base station to obtain a least squarestandard form; constructing an L1 norm expression form, and converting the L1 norm expression form into an L1 norm expression form with additional equation linear constraints; giving an iterative process of alternating direction method solving to obtain a positioning solution of the mobile station; substituting the positioning solution into the L1 norm expression to obtain an absolute value residual error, and if the median ratio of the maximum absolute value residual error to the absolute value residual error is greater than a preset value, outputting the positioning solution of the mobile station; otherwise, adopting a basic positioning equation between the mobile station corresponding to the next distance observation value and the base station as a reference equation. According to theinvention, the influence of non-line-of-sight errors on the positioning precision can be reduced.

The invention relates to an Internet forum page clustering method based on a website structure and equipment. The method relates to the field of data mining, and is provided by aiming at solving the large-scale webpage classification problem. The method comprises the following steps that partial samples are extracted from a website totality; each website is subjected to structure division by utilizing the characteristic of Internet forum website high structuralization; structure vectors are constructed; the distance between the structure vectors is evaluated by using a distance function provided by the invention; next, the structure vectors of the samples are subjected to clustering analysis by a density peak value clustering method; the feature structure of each cluster is extracted; a resolver for describing all sample websites in the clusters is constructed; and the resolver is used for resolving and classifying the rest websites in the totality. Experiments prove that the method has high accuracy and execution efficiency.

The invention discloses a topology dynamic identification method for a medium-voltage power distribution network. The method comprises the following steps: s1, acquiring upstream voltage, downstream voltage and time data of a power distribution terminal acquisition line; s2, by the power distribution terminal, broadcasting the attribute information of the power distribution terminal to the outsidethrough the network; s3, by the power distribution terminal, acquiring the attribute information of other power distribution terminals in the network through the network; s4, by the power distribution terminal, dynamically generating power grid topology information associated with equipment of the power distribution terminal according to broadcast information published by other power distributionterminals; s5, by the power distribution terminal, comparing the acquired attribute information of the power distribution terminal with the attribute information of the power distribution terminal; and s6, forming a topology network. The problem that topology information changes due to updating of topology information of a medium-voltage power distribution network and needs to be updated again manually is solved, so that a power distribution terminal can report, store and call the topology information of the power distribution network in a self-adaptive mode.

The invention discloses an oil storing barrel sealing machine of a motorcycle rear shock absorber. The oil storing barrel sealing machine comprises a rack (9), a motor (1), a sealing head assembly (2) and a sealing idler wheel (3), wherein the motor (1) is mounted on the rack (9); an output end of the motor (1) is connected with the sealing head assembly (2); and the sealing head assembly (2) is provided with three grabbing and clamping discs and the three grabbing and clamping discs are connected the sealing idler wheel (3). The oil storing barrel sealing machine of the motorcycle rear shock absorber has high sealing precision, does not have an oil leaking phenomenon, has a simple structure, and is economical and practical.

A pneumatic shock wave therapeutic apparatus relates to the field of medical devices. The present invention aims to solve the existing problems of how to ensure deep treatment and reduce harm, single treatment method and unsatisfactory treatment effect. A pneumatic shock wave therapy instrument, comprising a tube body, a handle body, a slider moving member, an impact device and a compressed air introduction device; one end of the tube body is inserted into the handle body, and the inside of the tube body is provided with a slider moving Components, an impact device and a compressed air introduction device; the compressed air introduction device and the impact device in an annular array are respectively arranged at two ends of the pipe body, and one end of the slider moving component is connected with the impact device. The present invention solves the existing problems of how to ensure deep treatment and reduce harm, single treatment method and unsatisfactory treatment effect, improves the treatment effect in the deep treatment area, changes the treatment method, and reduces pain The feeling guarantees the therapeutic effect again.