ATM Transaction Interface Recognition for Custom-Free Instruction Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional ATMs require significant labor to create function-specific instruction information due to varying customer specifications and apparatus arrangements, necessitating customized programming for each bank.
Innovation Solution
A transaction processing system that includes an acquisition unit to capture image data from a display, an extraction unit to identify characters or graphics, and a generation unit to generate processing instructions based on this data, allowing for adaptable operation across different ATM configurations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If conventional ATM systems use customized programming for each bank's specific interface specifications, then the system can accommodate varying customer requirements and apparatus arrangements, but the labor and complexity required to create and maintain these custom functions increases significantly
Solution Approach 1:
The system captures images of the display screen showing the customer interface, and uses OCR (optical character recognition) to extract text and information from these images. This copying approach allows the system to adapt to different interface designs without requiring custom programming for each variant, as the interface content is read directly from the captured images rather than being hard-coded
Solution Approach 2:
The patent replaces traditional mechanical programming approaches with image processing and AI-based recognition systems. Instead of manually configuring instruction information based on interface specifications, the system uses computer vision to automatically detect and interpret display content, generating appropriate instructions through pattern recognition rather than pre-programmed logic
2Manufacturing precision
If conventional ATM systems require manual creation of instruction information for each interface specification, then precise control over transaction processing is achieved, but the time and labor required for setup and maintenance increases
Solution Approach 1:
The system performs self-configuration by automatically capturing display images, extracting interface content through OCR, and generating transaction instructions without human intervention. The AI model learns from the captured images and autonomously adapts to different interface specifications, eliminating the need for manual programming while maintaining precise control over transaction processing
Solution Approach 2:
The system performs preliminary image capture and analysis during the setup phase, storing extracted interface information for future use. By pre-processing and storing the interface characteristics, the system eliminates the need for repeated manual configuration when dealing with the same interface type, significantly reducing setup time while maintaining processing precision
Data Source
Figure 1
Figure 2
Figure 3~4
AI summary
An information processing apparatus that can easily cope with the specification of the apparatus that differs for each customer, or a change in the specification of the apparatus is provided. A transaction processing system of the present disclosure includes an acquisition unit for acquiring image data of an image displayed on a display apparatus for displaying a content of a transaction, an extraction unit for extracting a character or graphic displayed on the display apparatus from the image data, a generation unit for generating instruction information corresponding to the operation performed on the input apparatus by the user from the changed character or graphic, and a communication unit for transmitting the instruction information to a money processing apparatus.