Image recognition device and data registration method for image recognition device
An image recognition device and data technology, applied in image data processing, image analysis, image enhancement and other directions, can solve the problems of no specific disclosure priority, unclear database update, no automatic database update proposed, etc., to achieve good recognition accuracy, The effect of increased diversity
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
no. 1 approach ＞
[0039] (system structure)
[0040] figure 1 It is a diagram schematically showing the functional configuration of the face matching system (image recognition device) according to the first embodiment of the present invention. The face verification system 1 is a device for verifying or identifying a person to be verified using a face image. Identity verification refers to the process of confirming whether the person to be checked is the person (one-to-one verification), and personal identification refers to the process of determining whether the person to be checked is one of the registrants registered in the database (one-to-many) check). These face verification technologies can be used in various applications such as security devices for electronic devices (including computers, smartphones, and tablet terminals), surveillance systems for detecting intruders, and access control systems for room entry and exit management or door lock control. Various uses.
[0041] Such as ...
no. 2 approach ＞
[0065] In the first embodiment, a structure is adopted in which new data similar to registered data is not added at all. In the second embodiment, a structure is adopted in which even new data similar to registered data is added. Registered structures are added until the number of registered data reaches the upper limit. This is because the greater the number of feature data, the higher the recognition accuracy can be expected compared to the smaller the number of feature data. The configuration other than that is the same as that of the first embodiment, so only the characteristic parts of the second embodiment will be described below.
[0066] Figure 6 It is a flowchart showing the flow of data registration processing in the face matching system of the second embodiment. The processing of steps S40 to S47 is completely the same as that of the first embodiment.
[0067] Next, the registration unit 15 checks whether or not the number of registered data of the registered fa...
no. 3 approach ＞
[0071] In the first embodiment, the similarity between newly created data and registered data is evaluated by the score between the two, while in the third embodiment, the similarity between newly created data and registered data is It is judged by evaluating the correlation (similarity) with the scores of other registered data.
[0072] The basic configuration and processing are the same as those of the first embodiment or the second embodiment, but Figure 4 or Figure 6 Step S45 (calculate score), step S48 (judgment of similarity between newly created data and registered data), and step S51 (select registered data to be replaced) in the flow of the process are different. Below, refer to Figure 7 , to illustrate the contents of these three processes.
[0073] In step S45 , the face collating unit 14 calculates scores for all combinations of newly created data and registered data. This score is the same index as that used for judging whether or not the face is the same f...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More