Image Identifier Detection Using Binarization and Connected Domain Calibration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current object detection methods, such as the sliding window approach, are computationally intensive and inefficient for recognizing specified identifiers on objects like bank cards, particularly when dealing with fixed structures, leading to longer detection times and increased complexity.
Innovation Solution
A method that involves retrieving and binarizing images from a predetermined area, calibrating connected domains, and comparing image features to determine the location of specified identifiers, reducing the detection area and computation complexity by using a virtual frame and standard image templates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If conventional object detection methods like sliding window are used, then detection can be performed on general objects, but computation time and complexity increase significantly
Solution Approach 1:
The patent segments the detection process into distinct phases: first detecting the bank card boundary using edge detection, then detecting identifier objects within the confined card area. This segmentation reduces the search space from the entire image to specific regions, dramatically reducing computation time while maintaining detection capability.
Solution Approach 2:
The patent performs preliminary actions by first detecting the bank card boundary and establishing a coordinate system before proceeding to identifier object detection. This preliminary boundary detection creates a constrained search space and provides reference frames for subsequent detection steps, reducing overall computation time.
2Reliability
If the entire image is searched for identifier objects, then no detection area is missed, but computation volume and complexity increase
Solution Approach 1:
The patent applies local quality by focusing computational resources on the specific region where identifier objects are expected to appear (within the bank card boundary). Instead of uniformly searching the entire image, the method concentrates detection efforts on the card area, reducing computation complexity while maintaining reliability for identifier detection.
Solution Approach 2:
The detection space is segmented into the bank card region and the surrounding area. By restricting identifier object detection to only within the detected card boundaries, the patent reduces the search space and computation complexity while ensuring that all relevant identifier objects are captured.
3Measurement precision
If affine transformation is applied to correct image perspective, then detection accuracy improves, but computation time increases
Solution Approach 1:
The patent performs affine transformation as a preliminary step after detecting the bank card boundary but before detailed identifier object detection. By establishing the transformed coordinate system in advance, the method ensures accurate detection throughout the process without requiring repeated transformations, optimizing the balance between accuracy and computation time.
Data Source
AI summary
Embodiments of the present application relate to a method, apparatus, and system for detecting a specified image identifier. The method includes retrieving a target image to be detected from a predetermined area, binarizing the target image to be detected to obtain a target binary image to be detected, calibrating connected domains of the target binary image to be detected, successively retrieving image features of candidate connected domains, and comparing the image features corresponding to the candidate connected domains to image features of a standard specified identifier image, wherein the candidate connected domains are determined based at least in part on the calibration of the connected domains, and determining a candidate connected domain as the location of the specified identifier image based at least in part on the comparison of the image features corresponding to the candidate connected domains to image features of the standard specified identifier image.