Patents
Literature
Hiro is an intelligent assistant for R&D personnel, combined with Patent DNA, to facilitate innovative research.
Hiro

10 results about "Multiple frame" patented technology

Face recognition method, device, computer device and storage medium

PendingCN109241868AEnsure safetyImprove recognition efficiencyCharacter and pattern recognitionFeature vectorImage extraction
The invention discloses a face recognition method, a device, a computer device and a storage medium. The face recognition method comprises the following steps: obtaining an image to be recognized, extracting a feature vector to be recognized according to the image to be recognized; acquiring a feature similarity between the feature vector to be identified and each reference feature vector; takingthe user identifier corresponding to the reference feature vector with the highest feature similarity as the target user identifier; obtaining a target region in the image to be recognized according to the target user identifier, and extracting a target feature vector according to the target region; calculating the vector similarity between the target feature vector and the user-defined feature vector corresponding to the target user identification, and obtaining the recognition result according to the vector similarity. The face recognition method provided by the invention does not need multiple frames of images for comprehensive verification, does it need to train a complex neural network model in advance to realize, and can ensure the safety of the face recognition process and improve the recognition efficiency at the same time.
Owner:PING AN TECH (SHENZHEN) CO LTD

A face feature point rapid alignment method based on an AAM

InactiveCN106228113ANarrow searchReduce computationCharacter and pattern recognitionPattern recognitionMultiple frame
The invention discloses a face feature point rapid alignment method based on an AAM, and belongs to the technical field of mode identification. The face feature point rapid alignment method of the invention comprises the steps of performing face feature point calibrating on an input face image so as to use the input face image as a training sample of a contour model; performing contour model training on the training sample by using an AAM active contour model method to obtain the contour model; inputting a continuous multi-frame video image to be processed; if the current frame is a starting frame, performing AAM search matching on the whole image to obtain a face feature point alignment result of the current frame and contour model; and if the current frame is not the starting frame, the face feature point alignment result of the previous frame is regarded as the starting position, and AMM searching and matching is carried out on the current frame to obtain the face feature point alignment result of the current frame and the contour model. The face feature point rapid alignment method of the present invention is applied to face recognition processing, the iteration initial position of the current frame is determined by the alignment structure of the previous frame, so that the search range and the number of iterations are reduced and the time consumption in the face feature point alignment is greatly reduced.
Owner:UNIV OF ELECTRONICS SCI & TECH OF CHINA

Down-the-hole radial gate with multiple frames and multiple trunnions and two lifting pad eyes

InactiveCN102979063ASmall opening and closing forceReduce capacityBarrages/weirsHydraulic cylinderMultiple frame
The invention discloses a down-the-hole radial gate with multiple frames and multiple trunnions and two lifting pad eyes. The down-the-hole radial gate with the multiple frames and the multiple trunnions and the two lifting pad eyes is characterized in that a multiple frame structure is adopted in the door body structure of a gate (1), and the gate (1) is hinged with a group of trunnions (4) installed on a breast wall (3) of a gate wall (2). The down-the-hole radial gate with the multiple frames and the multiple trunnions and the two lifting pad eyes reduces loads distributed to each main beam and each support arm through the multiple frame structure, shares huge concentrated water pressure of a high water head of a drainage port to the gate through the plurality of trunnions, and the loads distributed to each trunnion are greatly reduced so as to reduce the difficulties of the design, manufacture, welding, transportation and installation of the gate and the casting of the trunnions. The radial gate is provided with the two lifting pad eyes, is operated by adopting a double-cylinder swinging type or a suspension type hydraulic cylinder, the volume of single cylinder hydraulic cylinder is reduced by half so as to reduce the difficulties of the design, manufacture and installation of the hydraulic cylinder. The rotating center of the gate coincides with the circle center of an arc-shaped water-facing surface, and the total water pressure acting on the arc-shaped water-facing surface is enabled to always pass through the rotating center of the trunnions so as to reduce the force of the opening and closing of the gate.
Owner:CHINA POWER CONSRTUCTION GRP GUIYANG SURVEY & DESIGN INST CO LTD

Baby carriage frame welding positioning device

The invention discloses a baby carriage frame welding positioning device, which comprises a base. The two ends of the base are provided with handles. The invention solves the problem of welding and positioning of the frame parts of the baby carriage, and has simple design, reasonable structure, convenient installation of the frame parts, accurate positioning, convenient and firm welding.
Owner:ANHUI SHUBEILE CHILDREN THINGS

Living body detection and auxiliary living body detection method, device and equipment

PendingCN112215069ASolve the problem of no detection abilitySpoof detectionMedicineMultiple frame
The embodiment of the invention discloses a living body detection and auxiliary living body detection method, device and equipment, the living body detection method can receive a to-be-detected videoshot and uploaded by terminal equipment, and in the process of shooting the to-be-detected video by the terminal equipment, a vibration motor of the terminal equipment is in an on state; sequentiallyselecting multiple frames of images from the to-be-detected video, and determining the jitter degree of the to-be-detected video in the shooting process based on the multiple frames of images; and judging whether the to-be-detected video is a real shot living video based on the jitter degree.
Owner:ALIPAY LABS (SINGAPORE) PTE LTD
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products