CAPTCHA Image Segmentation for Bot-Proof Access Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current CAPTCHA systems become too complex for humans to recognize due to image distortion, while AI-powered computers can still identify them, striking a balance between ease for humans and difficulty for computers is challenging.
Innovation Solution
Segmenting a CAPTCHA image into multiple parts that utilize human persistence of vision to display a complete image, making it recognizable for users while being unrecognizable for computers, even if they capture screenshots of the parts.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If CAPTCHA images are distorted to prevent computer recognition, then security against bots is improved, but human recognizability deteriorates
Solution Approach 1:
The CAPTCHA image is divided into multiple segments that are displayed sequentially at different positions on the screen. Each segment contains only a portion of the original image, making it unrecognizable when viewed individually or captured via screenshot. When all segments are combined in the correct sequence, they reconstruct the complete recognizable image for human users who can track the progression across segments.
2Reliability
If CAPTCHA complexity is increased to defeat AI recognition, then security is improved, but user experience deteriorates
Solution Approach 1:
The CAPTCHA system transitions from a static image to a dynamic sequence of image segments displayed over time. The segments are presented in a time-based progression where each segment appears temporarily at a specific position before transitioning to the next segment. This dynamic presentation creates a time-sensitive challenge that defeats static AI analysis while remaining intuitive for human users who can process temporal sequences naturally.
3Ease of operation
If simple CAPTCHA patterns are used, then ease of recognition for humans is improved, but vulnerability to computer pattern recognition increases
Solution Approach 1:
The solution adds temporal and spatial dimensions to the traditional CAPTCHA image. Instead of a single static 2D image that can be analyzed by pattern recognition algorithms, the system presents multiple 2D segments across different spatial positions and time points. This transformation from a single-state image to a multi-state temporal sequence creates a higher-dimensional challenge that requires processing across multiple dimensions, defeating simple pattern matching while maintaining human recognizability through sequential assembly.
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
Ensures that humans can easily identify the original image without distortion, while computers cannot reconstruct the image from the segmented parts, enhancing security by leveraging human visual perception differences.
Implementation Method 1
Segmenting a CAPTCHA image into multiple parts that utilize human persistence of vision to display a complete image
Data Source
AI summary
Determining a group of figures for use in a vision test to distinguish computers from humans. An image is obtained and segmented into a plurality of parts. Based on the plurality of parts, a group of figures is determined to enable the group of figures to be displayed at a certain rate for a user to recognize the image.