177results about How to "Expand the scope of recognition" patented technology

The invention relates to high optical spectral image technique without harm to agricultural products. It can reflect the appearance of the agriculture products like color, shape, texture, dimension, scar and son on, and internal features like hardness, protein content and connected with knowledge base and experience of experts to make judgment. It can make quick, accurate, timely judgment of products, controlling the overall production with guarantee of the agriculture quality.

The invention discloses an intelligent conversation device and a feedback intelligent voice control system and method. The feedback intelligent voice control system comprises a voice input module, a voice recognition module, a data processing module, a local instruction library and an MCU. The local instruction library is used for storing text instructions and corresponding control instructions; the voice input module is used for collecting voice information input by a user; the voice recognition module is used for converting the voice information into text information; the data processing module is used for judging whether or not the local instruction library contains a text instruction identical with the text information, if not, fuzzy matching is conducted on the text information and the text instructions in the local instruction library, and if the text information is successfully matched with the target text instruction, a target control instruction corresponding to the target text instruction is transmitted to the MCU; the MCU is used for executing operation according to the target control instruction. By means of the intelligent conversation device and the feedback intelligent voice control system and method, the real intention of the user is automatically predicted in the intelligent man-machine conversation process, the user is rapidly and accurately responded to, the voice recognition rate is increased, and the voice recognition range is widened.

An ultrasound apparatus includes a touch screen configured to display, on an ultrasound image, a touch recognition region of an object used as a measurement mark; and a controller configured to move the object and the touch recognition region, in response to an input for touching and dragging the touch recognition region, to detect, from a portion of the ultrasound image which corresponds to the touch recognition region, a line formed by connecting points at which a brightness variation of a pixel is greater than a threshold value, and to move the object to a position of the detected line by using coordinates of the detected line.

Disclosed herein are complexes and compounds that pass through cellular barriers to deliver compounds into, through and out of cells, and methods of producing and using such complexes and compounds. The complexes and compounds of the invention comprise a biologically active portion and a targeting element directed to a ligand that confers transcellular, transcytotic or paracellular transporting properties to an agent specifically bound to the ligand. Also disclosed are complexes and compounds that comprise two or more targeting elements directed to a ligand that confers transcellular, transcytotic or paracellular transporting properties to an agent specifically bound to the ligand. Preferred ligands include but are not limited to the stalk of pIgR, a pIgR domain, an amino acid sequence that is conserved among pIgR's from different animals, and one of several regions of pIgR defined herein.

The invention discloses an intelligent wine cabinet system and an information processing method thereof. The intelligent wine cabinet system comprises an application server, and comprises an intelligent wine cabinet, a network server, a database server and a remote operating end which are all connected with the application server. A RFID (Radio Frequency Identification) module is used in the intelligent wine cabinet. The RFID module can identify the leaving and entering of the wine but also can identify the individual information of the wine. An information transmission interface is connected with the application server through a link, and the application server can perform information interaction with the database server, the network server and the remote operating end. With the intelligent wine cabinet system and the information processing method thereof, users can manage respective wine cabinets more conveniently and quickly, and simultaneously, the users can perform operations, such as management, query and the like, on respective wine cabinets via remote control.

Images of farm animal products taken by CCD camera through image acquisition card are transferred to computer. Under certain condition (temperature, humidity and flow rate), smells emitted from farm animal products through mini pump with no oil are sucked into reaction chamber, where signals are generated by array of gas sensor. The said signals through conditioning circuit, A / D acquisition card are transferred to computer. Computer for simulating human brain to treat vision data and smell data carries out syncretizing and pattern recognition processing. Combining with experts' knowledge in knowledge base as well as experiences, computer makes judgment for qualities of garden spgarden stuff, meat and corns etc. synthetically.

A method for automatic classification separately collection and automatic transportation of the solid waste is designed to replace the manual door to door collection of the waste by the computer, and getting the waste for valuable recycle and reuse comprises of putting the waste into the automatic waste collectors set at the source of the waste producing and connected with pipeline, identifying the waste by the main board (21), classifying the waste by the selector (6), depositing the waste in the containers (1), (2) and (3), fluidizing the wastes by rolling fluidizing assembly (15), under the monitoring of the supervisory control center, separately transporting the wastes into the correspond waste recycle center, and getting the solid waste automation.

Embodiments of the present specification disclose a data processing method, apparatus and device for insurance fraud identification, and a server. The method may comprise: acquiring relationship association data of a target population, wherein the target population comprises a set of selected applied claimants and insurants; constructing multi-degree relationship network map data for the target population based on the relationship association data; clustering the multi-degree relationship network map data by using an unsupervised learning algorithm to determine a classified communities of thetarget population and corresponding community information; and calculating community characteristics of the classified communities based on the community information, and marking the classified community that meets the insurance fraud decision-making condition as a fraud group. According to the embodiments of the present specification, a group of fraudsters can be effectively and efficiently identified from a multi-degree relationship.

The embodiment of the invention provides a two-dimension code recognition method and device and relates to the technical field of image recognition. The two-dimension code recognition method and device can recognize distorted two-dimension codes and improve the recognition rate of the distorted two-dimension codes. The method comprises the steps that the area where a two-dimension code is located and three gravity center points corresponding to three position detection images of the two-dimension code one to one are determined in the two-dimension code to be recognized; key points are determined according to the area where the two-dimension code is located and the three gravity center points; four edge fitted curves are obtained according to four corner points of the area where the two-dimension code is located and the outline of the area where the two-dimension code is located; the mapping relation between the image of the area where the two-dimension code is located and a corrected image is determined according to the coordinates of the key points in the two-dimension code image, the four edge fitted curves and the corrected image of the two-dimension code; the image of the area where the two-dimension code is located is corrected according to the mapping relation, and the information of the corrected image of the area where the two-dimension code is located is obtained. The two-dimension code recognition method is applied to a two-dimension code recognition system.

The invention discloses an optical fiber intrusion signal recognition method based on a support vector machine, so as to solve the problem of how to accurately and effectively recognize the type of an optical fiber intrusion signal in a phase sensitive optical time domain reflection mechanism. The realization method comprises steps: time-frequency domain features of an optical fiber intrusion signal sample are firstly extracted to serve as a training sample; a kernel function is then used for mapping the sample to high-dimensional space; sequential minimization algorithm is then used for training a support vector machine classifier and mesh optimization and cross validation are used for acquiring the optimal classifier parameters; and finally, the feature vector of a to-be-classified optical fiber intrusion signal is inputted to the classifier, and a corresponding classification result is obtained according to output of the classifier. According to the method of combining the support vector machine and digital features of the optical fiber intrusion signal, the intrusion type can be accurately recognized.

The invention discloses an obstacle avoidance method and an obstacle avoidance system of a mechanical arm. The obstacle avoidance method of the robotic arm includes the steps of: collecting image information of the obstacle, and collecting distance information between the robotic arm and the obstacle; analyzing the image information of the obstacle and the distance information to determine whether a new obstacle appears on the robot within the preset range of the robotic arm; if the new obstacle appears within the preset range of the robotic arm, re-plan the operating path of the robotic arm and control the movement of the robotic arm according to the re-planned operating path; if there is no new obstacle When an obstacle appears within the preset range of the robot arm, the robot arm is controlled to move on the original running path. The obstacle avoidance method of the above-mentioned mechanical arm judges whether there is a new obstacle by image information and distance information, which can solve the shortcomings of the blind area in the simple visual obstacle recognition method, expand the obstacle recognition range, and then effectively improve the obstacle recognition. The safety and efficiency of the robot reduce the risk of the robot colliding with obstacles.

The invention relates to a dairy cow identification method and a system. The method comprises the steps of acquiring the real-time image of a to-be-identified dairy cow; matching the real-time image with images pre-established in an image template library to acquire a matching image matched with the real-time image; determining the identification number of the to-be-identified dairy cow according to the identifier of the matching image. According to the invention, the identification range of dairy cows is expanded. Meanwhile, no electronic tag needs to be installed on the bodies of dairy cows.

Disclosed herein are multimeric molecular complexes and compounds that are multivalent, i.e., they have two or more targeting elements directed to a ligand that confers paracellular transporting properties and / or transcytotic properties to complexes and compounds to which it is bound. The complexes and compounds have properties that are different from the properties of monomers, complexes and compounds having only one targeting element directed to a paracellular and / or transcytotic ligand. The complexes and compounds of the invention undergo endocytosis, transcytosis and exocytosis; following endocytosis, the complexes or compounds may be transported into the cytosol or an organelle of a cell. In polarized cells, transcytosis can proceed in a “forward” or “reverse” direction. Reverse transcytosis is used for the non-invasive delivery of biologically active agents from the lumen of, e.g., the gastrointestinal tract or the airways of lungs, to the circulatory system. The complexes and compounds are incorporated in various compositions and medical devices suitable for medicinal or veterinary use.

The invention provides a PSCA (prostate stem cell antigen) and PD-L1 targeted CAR based on OCTS (one CAR with two ScFvs)-CAR (chimeric antigen receptor), an encoding gene, an OCTS-CAR-T recombinant expression vector as well as a construction method and an application of the OCTS-CAR-T recombinant expression vector. The PSCA and PD-L1 targeted CAR comprises a CD8 leader membrane receptor signal peptide, a double-antigen binding region, a CD8 Hinge chimeric receptor linker, a CD8 Transmembrane chimeric receptor transmembrane region, a CD28 chimeric receptor co-stimulator, an OX40 chimeric receptor co-stimulator and a TCR chimeric receptor T-cell activation domain which are connected in series sequentially, wherein the double-antigen binding region comprises heavy chains VH and light chains VL of PSCA and PDL1 single-chain antibodies, Inner-Linker in antibodies and Inter-Linker among the single-chain antibodies, which are in series connection or turn connection. Besides, the invention provides the encoding gene of the PSCA and PD-L1 targeted CAR, the recombinant expression vector as well as the construction method and the application of the recombinant expression vector.

The embodiment of the invention discloses a safety early warning method in a kitchen and a range system. The method comprises steps: whether a person enters a preset danger area in the kitchen is detected in real time; when a person is detected to enter the danger area, whether the person is a kid is detected; and when the person is detected to be a kid, safety early warning is carried out according to a preset safety early warning scheme. Through the scheme disclosed in the embodiment of the invention, safety early warning is carried out when a kid enters the danger area, the kid can be effectively prevented against a further dangerous behavior, and the safety of the kid after entering the kitchen is improved.

An embodiment of the invention discloses an insurance fraud identification data processing method, device, equipment and server. Multi-scale relationship network graph data of crowds are built based on multi-scale relationship association data of insurance applicants and the insured, a relationship network among people can be more deeply mined, identification efficiency is improved, and identification range is widened. Besides, a supervised learning model is jointly built according to characteristic data of insurance fraud people and used for learning relationship network characteristics and characteristics of the insurance fraud people. Ganged insurance fraud people have obvious multi-scale relationship characteristics on the relationship network, the characteristics of the insurance fraud people frequently show up similarity, so that the insurance fraud people can be more effectively and efficiently identified by the method, and identification processing efficiency is improved.

The invention discloses a traffic lane detection method based on an equidistant curve modle. The method can be used for recognizing multiple traffic lanes in an image obtained by an on-board camera, is large in recognition range, ensures that the recognition result is not confined to simple curves of straight lines, arcs and the like, also can be used for recognizing traffic lanes with variable curvature radius, recognizing the type of each traffic lane and discriminating real lines from dashed lines and is high in recognition accuracy. The method disclosed by the invention comprises the following steps: performing reversed projection conversion on a panoramic image of the front part of a vehicle, wherein the panoramic image is obtained by the on-board camera; performing filtration and binarization pretreatment on a ground overhead view; performing Hough conversion on the image subjected binarization pretreatment; recognizing the traffic lanes and types thereof in a Hough space.

The invention provides an anti-collision sensor and a welding robot, and relates to the technical field of sensors. The anti-collision sensor is connected with a movable work end. When the movable work end bears collision force, a transmission piece in the anti-collision sensor transmits the collision force to a sliding piece so that the sliding piece can slide relative to a shell. A control mechanism is used for stopping movement of the movable work end when a sensing mechanism senses that the displacement amount, relative to the shell, of the sliding piece reaches a preset value. An elastic piece is arranged between the sliding piece and the inner wall of the shell. After external force of the movable work end is removed, the elastic piece drives the sliding piece to slide to the balance position. The transmission piece comprises a transmission plate and a transmission shaft. A conversion assembly is arranged between the transmission plate and the inner wall of the shell so that the transmission plate can move in the sliding direction after the transmission shaft bears force not in the sliding direction and the anti-collision sensor can recognize collision force in various directions.

The invention discloses an automatic filing system and method for court electronic files. A computer automated processing technique is used, the image feature values of the electronic file titles arelearned, trained and recognized through the artificial intelligence technology and matched with the preset feature library, the functions of automatic recognition and automatic filing of the electronic files are achieved, new types are continuously learned through a learning training model to obtain new classification feature values, and therefore the feature library is continuously updated to enrich the classification types. According to the invention, the computer and artificial intelligence are combined, and manual intervention is taken as assistance to replace an original manual processingmode to archive and classify the electronic files of the court, so that the working efficiency and the archiving accuracy are effectively improved; and through the learning training model, the computer continuously learns to adapt to more types of files, so that the filing classification is further perfected, and the recognition range of the system and method is expanded.

The invention relates to wide-recognition-range monitoring equipment based on the face recognition technology. The wide-recognition-range monitoring equipment comprises a sliding rail, a movable block, an adjusting mechanism, a camera and a clearing mechanism. The adjusting mechanism comprises a rotary block, a rotary shaft, a movable assembly, a rotary assembly, a swing assembly and two supporting plates. The rotary assembly comprises a first motor, a conveying belt, two driving wheels, two driving units and two driven units. Each driving unit comprises a connecting shaft, a first gear, a first gear rack and a limiting unit. According to the wide-recognition-range monitoring equipment based on the face recognition technology, the angle of the camera is adjusted through the adjusting mechanism, and therefore the monitoring recognition range is widened. The practicality of the monitoring equipment can be improved. Dust accumulated on a lens of the camera is cleared away through the clearing mechanism, it is avoided that the cleanliness degree of the camera is reduced as the use time of the camera is increased, and therefore the recognition precision of the monitoring equipment is improved.

The invention discloses a positioning device for an unmanned line-tracking robot. The positioning module comprises a power supply, a microcontroller, n radio frequency readers, an RS232 communication interface and a host computer, wherein serial ports of the n radio frequency readers are connected with the serial ports of the microcontroller respectively, the power supply supplies power for the microcontroller, and an asynchronous serial port of the microcontroller is connected with the RS232 interface of the host computer through the RS communication interface. The positioning device can realize the precise positioning of the robot. Meanwhile, the invention further discloses a positioning method of the positioning device, and the positioning method is simple and easily operated, and can realize the precise positioning of the robot during the travelling of the robot.

The invention relates to high optical spectral image technique without harm to agricultural products. It can reflect the appearance of the agriculture products like color, shape, texture, dimension, scar and son on, and internal features like hardness, protein content and connected with knowledge base and experience of experts to make judgment. It can make quick, accurate, timely judgment of products, controlling the overall production with guarantee of the agriculture quality.

The invention discloses a peer-to-peer network application traffic identification method and relates to the application of computer peer-to-peer network. The identification method is mainly applied in an open network environment. The invention adopts a traffic identification method combining deep packet scanning and deep stream scanning to identify the present increasing traffic of peer-to-peer network application, wherein deep stream detection is to mainly compare collected network traffic data with feature codes in a feature library so as to distinguish whether network traffic includes the traffic of peer-to-peer network application; and subsequently the deep stream detection distinguishes whether the network traffic includes the traffic of peer-to-peer network application based on the symmetry of the upstream and downstream traffic of the network traffic data and the linking number between peer nodes and other peer nodes. The identification method has the advantages of effectively expanding the identification range of traffic of peer-to-peer network application, boosting the accuracy rate of identifying the traffic of peer-to-peer network application and consequently guaranteeing the healthy, standardized and ordered development of peer-to-peer network application.

The invention discloses a lymphoblastic leukemia CAR-T (Chimeric Antigen Receptor-T Cell Immunotherapy) therapy carrier based on an OCTS (One CAR with Two SeFvs) technology. The lymphoblastic leukemia CAR-T therapy carrier comprises a lentivirus skeleton plasma, a human EF1alpha promoter, an OCTS chimeric receptor structural domain and an IL6R single-chain antibody, wherein the OCTS chimeric receptor structural domain comprises a CD8 leader chimeric receptor signal peptide and two groups of single-chain antibodies; the first group of single-chain antibodies is selected from any one of the following four groups of single-chain antibodies: a CD20 single-chain antibody light chain VL and a CD20 single-chain antibody heavy chain VH, a CD22 single-chain antibody light chain VL and a CD22 single-chain antibody heavy chain VH, a CD30 single-chain antibody light chain VL and a CD30 single-chain antibody heavy chain VH, and a CD123 single-chain antibody light chain VL and a CD123 single-chain antibody heavy chain VH; and the second group of the single-chain antibodies is a CD19 single-chain antibody light chain VL and a CD19 single-chain antibody heavy chain VH, an antibody Inner-Linker, a single-chain antibody Inter-Linker, a CD8-Hinge chimeric receptor linker, a CD8 Transmembrane chimeric receptor transmembrane zone, a TCR (T Cell Receptor) chimeric receptor T cell activation domain and a chimeric receptor co-stimulator zone. Besides, the invention discloses a constructing method for the carrier and application of the carrier to preparation of a drug for treating lymphoblastic leukemia.

The invention discloses an OCTS (One CAR (Chimeric Antigen Receptor) with two ScFvs (Single-chain variable Fragments)) technique based CAR-T (Chimeric Antigen Receptor-T cell immunotherapy) therapeutic vector for glioblastoma. The OCTS technique based CAR-T therapeutic vector comprises a lentiviral skeleton plasmid, a human EF1 [alpha] promoter (SEQ ID NO. 14), an OCTS chimeric receptor structural domain and a PDL1 single-chain antibody, wherein the OCTS chimeric receptor structural domain comprises a CD8 leader chimeric receptor signal peptide (SEQ ID NO. 15), a PDL1 single-chain antibody light chain VL (SEQ ID NO. 16), a PDL1 single-chain antibody heavy chain VH (SEQ ID NO. 17), an EGFRvIII single-chain antibody light chain VL (SEQ ID NO. 18), an EGFRvIII single-chain antibody heavy chain VH (SEQ ID NO. 19), an antibody Inner-Linker (SEQ ID NO. 20), a single-chain antibody Inter-Linker (SEQ ID NO. 21), a CD8 Hinge chimeric receptor linker (SEQ ID NO. 22), a CD8 Transmembrane chimeric receptor transmembrane domain (SEQ ID NO. 23), a TCR (T Cell Receptor) chimeric receptor T cell activation domain (SEQ ID NO. 26) and a chimeric receptor co-stimulator domain. In addition, the invention also discloses a construction method of the vector and application of the vector to the preparation of a medicine for treating the glioblastoma.

The invention discloses a website recognition method and device. The method comprises the steps: obtaining page image on a to-be-detected website; inputting a vector formed by the pixel values of a preset region in a webpage image into a pre-built target neural network, and obtaining an output result, wherein the target neural network is used for recognizing a phishing website; and recognizing whether the to-be-detected website is a phishing website or not according to an output result. The method solves a technical problem that a conventional website recognition technology is lower in accuracy.

The invention discloses an image recognition method, which comprises the following steps: obtaining a first image; Identifying the first image based on a preset first image recognition model to obtaina first recognition result; The first recognition result characterizes at least one first attribute represented by the first image and a confidence level corresponding to each of the at least one first attribute; At least accord to that first recognition result, it is determined whet the first image is a target class image. The invention also discloses an image recognition device and a computer-readable storage medium.

An infrared imaging sensor compatible with 2nd Generation Forward Looking Infrared (FLIR) Horizontal Technology Integration (HTI) B-Kit based sensors. In one example, the infrared imaging sensor includes a set of refractive opto-mechanical modules, including an afocal optical module, a receiver assembly, and backward- and forward-compatible electronics modules. The afocal optical module is configured to provide a plurality of different fields of view for the infrared imaging sensor. In one example, the sensor is configured for MWIR and LWIR imagery.

The invention relates to a method and a terminal for prompting ringtones according to calling numbers, wherein the method comprises the following steps: after a terminal receives a calling signal, acquiring a calling number from the number signal, wherein the calling signal includes phone incoming or message incoming; when the calling number does not belong to stored numbers in the terminal, playing a corresponding ringtone according to a mapping table of figures at preset positions in the calling number and preset characteristic digital ringtones. According to the invention, through carrying out matching and classification on figures at preset positions in a calling number which is not stored in the mobile terminal and characteristic figures, corresponding ringtones and information pictures and ringtones corresponding to the characteristic figures are respectively generated according to the classification, therefore, the recognition range of the calling number is expanded, and a defect that in the prior art, ringtones and information pictures and ringtones just can be generated according to calling numbers stored in the mobile terminal, but not generated through the classification of calling numbers which are not stored in the mobile terminal is overcome, thereby further improving the effect of user experience.