Biological feature recognition multi-modal fusion method and device, storage medium and equipment
A biometric identification and biometric technology, applied in the field of biometrics, can solve problems such as low quality of biometric samples, limitations of single-mode biometrics or reduced recognition accuracy, and inability to meet the requirements of high-security occasions.
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0034] The embodiment of the present invention provides a biometric feature recognition multimodal fusion method, such as figure 1 As shown, the method includes:
[0035] S100: Collect multiple types of biometric samples.
[0036] In this step, the types of biometric samples include but are not limited to: face, iris, fingerprint, finger vein, palm print, voice print, etc. For example: collecting face samples and iris samples, or collecting fingerprint samples and finger vein samples.
[0037] And each type of biometric sample can be a source, such as a face sample, because the same person has only one face, so there is only one source for his face sample; each type of biometric sample can also be a different source, such as The same person has two eyes, so their iris samples have two sources, and the same person has ten fingers, so their fingerprint samples and finger vein samples have ten sources each.
[0038] There is at least one source of each type of biometric sample...
example 1
[0138] Registration phase:
[0139] Step 1. Collect multiple types of biometric sample images, which include human face samples and iris samples.
[0140] In this step, the face image and the iris image can be collected simultaneously by the same camera, or the face image and the iris image can be collected separately by two cameras.
[0141] Step 2. Perform quality analysis on the face sample and the iris sample, and obtain the face sample quality score and the iris sample quality score.
[0142] Step 3. Extract and store the face features in the face sample and the iris feature in the iris sample respectively, to obtain a face template and an iris template.
[0143] Identification / verification phase:
[0144] Step 4. Execute steps 1-3 for the user to be tested to obtain the face features and iris features to be tested.
[0145] Step 5. Comparing the face feature to be tested with the face template to obtain a face comparison result, and comparing the iris feature to be te...
example 2
[0152] To simplify the description, this example omits the registration process and only retains the identification / authentication process.
[0153] Step 1. Collect multiple biometric sample images, the biometric sample images include face images and iris images.
[0154] In this embodiment, the face image and the iris image are collected by the same camera, and the face image and the iris image are located in the same image.
[0155] Step 2, perform quality analysis on the face image and iris image, and obtain the quality score of the biometric sample image.
[0156] Step 3. The preset fusion strategy in this example is feature data fusion and comparison score fusion. It is judged whether the quality score of the biometric sample image is greater than the feature data fusion threshold. If it is greater, face detection is performed on the biometric sample image. , extract face features, and at the same time locate the position of human eyes in the process of face detection, p...
PUM
Abstract
Description
Claims
Application Information
- R&D Engineer
- R&D Manager
- IP Professional
- Industry Leading Data Capabilities
- Powerful AI technology
- Patent DNA Extraction
Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.
© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com