Camera Autofocus Focal Distance for Photo Spoof Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for recognizing remade photographs in the context of identity authentication, such as those used in Taobao shop certifications, are complex and require additional hardware or sophisticated software, leading to inefficiencies in distinguishing between real and manipulated images.
Innovation Solution
The method employs an auto-focus function of a photographing device to determine whether a photographed object is a photograph or a real person by analyzing the focal distance after successful auto-focusing, without the need for additional hardware, using a simple software process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If fingerprint verification hardware is added to identify remade photographs, then identification accuracy is improved, but device complexity increases
Solution Approach 1:
The patent replaces the mechanical/fingerprint verification hardware with an optical analysis method. The system uses the camera's existing autofocus mechanism to detect focal distance, substituting complex hardware verification with a simpler optical measurement approach that leverages the camera's built-in capabilities.
Solution Approach 2:
The patent makes the camera system self-sufficient by using its own autofocus and focal distance measurement capabilities to detect photograph remaking. No external hardware is needed - the camera uses its inherent optical properties to identify whether it is capturing a real person or a photograph.
2Measurement precision
If video live detection with face detection algorithm is implemented, then identification capability is improved, but software complexity and processing time increase
Solution Approach 1:
The patent extracts the essential identification capability from complex software algorithms and reduces it to a simple focal distance measurement. Instead of implementing full face detection and action recognition software, the system isolates and uses only the focal distance parameter, which can be obtained from the existing autofocus mechanism.
Solution Approach 2:
The patent changes the identification parameter from complex multi-dimensional data (face features, actions, expressions) to a single scalar parameter (focal distance). This parameter change simplifies the software requirements while maintaining identification effectiveness, as focal distance naturally differs between real persons and photographs.
3Measurement precision
If additional hardware facilities are installed for photograph verification, then detection accuracy is improved, but manufacturing cost and device complexity increase
Solution Approach 1:
The patent makes the camera's autofocus system serve dual purposes: normal focusing for photography and focal distance measurement for photograph verification. This multi-functionality eliminates the need for dedicated verification hardware, reducing manufacturing costs while maintaining detection accuracy.
Solution Approach 2:
The system uses the camera's own autofocus mechanism to provide verification functionality, making the device self-sufficient. The camera measures focal distance using its built-in sensors and processing, eliminating the need for external verification hardware and reducing overall system cost.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach simplifies the recognition of remade photographs, reducing complexity and enhancing efficiency by leveraging the difference in focal distances between real and manipulated images, thereby improving the authentication process.
Implementation Method 1
activate a photographing device of a terminal to auto-focus a photographed object
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A method, an apparatus, a mobile terminal, and a camera for recognizing a photographed object are disclosed in the present application. The recognition method includes starting a photographing device of a terminal to auto-focus a photographed object in response to receiving a photographing instruction; obtaining a post-focusing focal distance of the photographing device after the photographing device successfully focuses on the photographed object; and determining whether an object type of the photographed object is a photograph based on the post-focusing focal distance. The present application resolves the technical problem of the high complexity associated with the schemes for recognizing a remade photograph in existing technologies.