Search

CN-121726012-B - Intelligent morning check equipment

CN121726012BCN 121726012 BCN121726012 BCN 121726012BCN-121726012-B

Abstract

The application relates to intelligent morning inspection equipment which comprises an equipment main body, a display screen arranged on the equipment main body, a first image sensor, a structural light sensor, an information acquisition table, a sealing cover, a second image sensor and a controller, wherein the sealing cover is arranged on the information acquisition table and is used for forming a closed space on the information acquisition table, the second image sensor and the controller are positioned in the closed space, the controller is used for carrying out data interaction with the first image sensor, the second image sensor and the structural light sensor and is used for carrying out identity recognition, living body authentication and abnormal area recognition in a hand image on a characteristic object, the first image sensor is used for acquiring facial information of the characteristic object, the structural light sensor is used for projecting dot matrix information to facial two-dimensional information of the characteristic object, and the second image sensor is used for acquiring the hand image of the characteristic object. The intelligent morning inspection equipment disclosed by the application combines a living body identification and local comparison detection mode to realize uniqueness of personnel identity and faster detection passing speed.

Inventors

  • PAN HAIJUN

Assignees

  • 君华高科集团有限公司

Dates

Publication Date
20260508
Application Date
20260224

Claims (8)

  1. 1. An intelligent morning check device, comprising: the device comprises a device main body (1), wherein a display screen (11) is arranged on the device main body (1); The first image sensor (2) and the structural light sensor (3) are both arranged on the equipment main body (1), the first image sensor (2) is used for collecting facial information of a characteristic object, and the structural light sensor (3) is used for projecting lattice information to facial two-dimensional information of the characteristic object; an information acquisition table (4) provided on the device main body (1); the sealing cover (5) is arranged on the information acquisition table (4), and the sealing cover (5) is used for forming a closed space on the information acquisition table (4); The second image sensor (6) is arranged on the information acquisition table (4) and positioned in the closed space, and the second image sensor (6) is used for acquiring hand images of the characteristic objects; The controller (7) is in data interaction with the first image sensor (2), the second image sensor (6) and the structure light sensor (3) and is used for carrying out identity recognition, living body authentication and abnormal region recognition on the characteristic object; the structured light sensor (3) projects lattice information to feature object face two-dimensional information, including: acquiring a true random number sequence from a clock and/or facial information of a feature object acquired by a first image sensor (2); generating a two-dimensional information projection lattice by using a true random number sequence, wherein the two-dimensional information projection lattice comprises a plurality of array areas, and each array area comprises MxN information elements; replacing at least one information element in each array area with an identity code to obtain lattice information; Projecting lattice information to the face of the feature object; when the characteristic object performs identity recognition and living body authentication, lattice information is projected to the face of the characteristic object for a plurality of times; And (5) re-acquiring a real random number sequence every time the two-dimensional information projects dot matrix information.
  2. 2. The intelligent morning check equipment of claim 1, further comprising: creating an image sequence attributed to the feature object; Sequentially acquiring facial images of the feature objects on a time sequence and judging imaging quality of the facial images; Marking the facial image of the characteristic object conforming to the imaging quality as a qualified image; determining the type of the qualified image, wherein the type comprises identification type qualified image and authentication type qualified image; Placing the qualified images into an image sequence, wherein the image sequence comprises at least one identification type qualified image and at least one authentication type qualified image; and when the number of the identification type qualified images or the authentication type qualified images in the image sequence meets the requirement, discarding the identification type qualified images or the authentication type qualified images continuously obtained on the time sequence.
  3. 3. The intelligent morning check equipment according to claim 1, wherein identifying an abnormal region in the hand image comprises: acquiring a contrast hand image according to the facial information of the feature object; comparing the acquired hand image with the comparison hand image to obtain a comparison result; dividing the hand image into areas according to the comparison result to obtain a normal area and an abnormal area; Wherein when an abnormal region occurs, an inspection image is generated using the hand image and the abnormal region.
  4. 4. The intelligent morning check equipment according to claim 3, wherein comparing the acquired hand image with the comparative hand image comprises: extracting fixed feature points of the hand image and fixed feature points of the contrast hand image by using a feature point matching algorithm; Fine-tuning the hand image by using the fixed feature points of the hand image and the fixed feature points of the contrast hand image to enable the hand image and the contrast hand image to coincide; Calculating pixel difference values of corresponding pixel points on the hand image and the contrast hand image to obtain a pixel difference value diagram; Generating a comparison result according to the pixel difference value graph; when the pixel difference value of the corresponding pixel point on the hand image and the contrast hand image is calculated, a plurality of monochromatic channels are used for processing the hand image and the contrast hand image, and a pixel difference value diagram belonging to each monochromatic channel is obtained.
  5. 5. The intelligent morning check equipment according to claim 4, further comprising checking the abnormal region after the abnormal region is obtained, and performing deletion processing through the checked abnormal region; Checking the abnormal region includes: determining the edge profile of the abnormal region; extracting analysis content from the hand image by using the edge contour of the abnormal region, and extracting reference content from the contrast hand image; Comparing the gray scale, texture and morphology of the analysis content and the reference content; And when any one of the gray comparison, the texture comparison and the morphological comparison does not pass, deleting the abnormal region, and otherwise, reserving the abnormal region.
  6. 6. The intelligent morning check equipment according to claim 5, wherein comparing the gray levels of the analysis content and the reference content comprises: Dividing analysis content and reference content into a plurality of cells respectively, wherein each cell comprises the same number of pixels; Calculating the pixel mean value of the cell; comparing the pixel mean values of adjacent cells and assigning values to the cells according to the comparison result; and calculating the similarity between the analysis content and the reference content and giving a gray comparison result.
  7. 7. The intelligent morning check equipment according to claim 5, wherein comparing textures of the analysis content and the reference content comprises: Dividing analysis content and reference content into a plurality of cells respectively, wherein each cell comprises the same number of pixels; Calculating the pixel mean value of the cell; determining a comparison line segment and calculating the difference value of adjacent cells on the comparison line segment; Generating a comparison number sequence according to the difference value, wherein the comparison number sequence is a difference value sequence or a secondary difference value sequence; Calculating the similarity of two comparison number sequences on the same comparison line segment, wherein the two comparison number sequences respectively belong to analysis content and reference content; and counting the similarity of the two comparison sequences on all the comparison line segments and giving out a texture comparison result.
  8. 8. The intelligent morning check equipment according to claim 5, wherein comparing the morphology comparison of the analysis content and the reference content comprises: obtaining a gray level histogram of the analysis content and a gray level histogram of the reference content; selecting an abnormal value point and a suspected abnormal value point adjacent to the abnormal value point on a gray level histogram of the analysis content; extracting on analysis content by using abnormal value points to obtain a first extraction result, and calculating the quantity value of the first extraction result; Respectively extracting the analysis content and the reference content by using suspected abnormal value points to obtain a second extraction result, and calculating the similarity of the second extraction result; and giving a morphological comparison result according to the quantity value of the first extraction result and the similarity of the second extraction result.

Description

Intelligent morning check equipment Technical Field The application relates to the technical field of safety inspection, in particular to intelligent morning inspection equipment. Background The breakfast morning inspection equipment is an intelligent terminal specially designed for screening the health and sanitation of kitchen personnel before going on duty in the breakfast, and aims to replace manual morning inspection, realize standardization, digitalization and traceability of detection, and the current detection range comprises identity verification, health screening, data management and control, linkage and the like. The morning check flow can be divided into two parts of identity recognition and security check, and in consideration of the working environment of the canteen, the current identity recognition is mostly carried out in a face recognition (non-contact) mode, and meanwhile, an image check mode is used for finding out problems (wounds, diseases, dirt residues and the like). The problem that mainly exists lies in using false information to punch a card, AI discernment is inaccurate and processing speed is slower, and false information to punch a card can lead to image inspection mode distortion, and the problem of AI discernment is then can directly influence the accuracy of judgement, causes unnecessary review work, and processing speed is slower and can lead to personnel backlog condition, because canteen morning is examined and is personnel concentrated scene, and processing speed is slower and can lead to personnel backlog, directly influences the development of follow-up work. Disclosure of Invention The application provides intelligent morning inspection equipment, which combines a living body identification and local comparison detection mode to realize uniqueness of personnel identity and faster detection passing speed. The above object of the present application is achieved by the following technical solutions: the application provides intelligent morning check equipment, which comprises: the main body of the device is provided with a plurality of air channels, a display screen is arranged on the equipment main body; The first image sensor is used for collecting the face information of the characteristic object, and the structural light sensor is used for projecting dot matrix information to the face two-dimensional information of the characteristic object; the information acquisition table is arranged on the equipment main body; the sealed cover is arranged on the information acquisition table and is used for forming a closed space on the information acquisition table; The second image sensor is arranged on the information acquisition table and positioned in the closed space, and is used for acquiring hand images of the characteristic objects; And the controller is in data interaction with the first image sensor, the second image sensor and the structural light sensor and is used for carrying out identity recognition, living body authentication and abnormal region recognition in the hand image on the characteristic object. In one possible implementation of the present application, the projecting lattice information by the structured light sensor onto the feature object face two-dimensional information includes: Acquiring a true random number sequence, wherein the true random number sequence is from a clock and/or facial information of a feature object acquired by a first image sensor; generating a two-dimensional information projection lattice by using a true random number sequence, wherein the two-dimensional information projection lattice comprises a plurality of array areas, and each array area comprises MxN information elements; replacing at least one information element in each array area with an identity code to obtain lattice information; The lattice information is projected to the face of the feature object. In one possible implementation manner of the application, when a characteristic object performs identity recognition and living body authentication, lattice information is projected to face two-dimensional information of the characteristic object for a plurality of times; And (5) re-acquiring a real random number sequence every time the two-dimensional information projects dot matrix information. In one possible implementation manner of the present application, the method further includes: creating an image sequence attributed to the feature object; Sequentially acquiring facial images of the feature objects on a time sequence and judging imaging quality of the facial images; Marking the facial image of the characteristic object conforming to the imaging quality as a qualified image; determining the type of the qualified image, wherein the type comprises identification type qualified image and authentication type qualified image; Placing the qualified images into an image sequence, wherein the image sequence comprises at least one identification type qualified image and