Non-overlapping Image Merging via Calibration Unit Coordinate Transformation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image merging methods struggle to achieve high precision and larger field of view in inspection processes, especially in environments with non-overlapping images and complex backgrounds like food production lines, where distinguishing characteristic information is challenging due to similar-sized pastes and stuffing.
Innovation Solution
An image merging method and device utilizing a calibration unit with multiple calibration devices and image capturing units to establish a conversion formula based on known and image characteristic information, allowing for the stitching of non-overlapping images into a composite image with higher precision and a larger field of view.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple image capturing devices with higher pixel density are used to enhance detection precision, then detection precision is improved, but the cost and device complexity increase
Solution Approach 1:
The imaging area is divided into multiple non-overlapping sub-areas, each captured by a separate image capturing device. Multiple images are then merged to form a composite image that provides high detection precision across the entire imaging area, avoiding the need for a single high-pixel-density device
Solution Approach 2:
Multiple images captured from different sub-areas are merged into a single composite image through coordinate transformation and pixel mapping. This merging process achieves high detection precision across the entire imaging area while using multiple standard-resolution devices instead of one high-resolution device
2Measurement precision
If the visual range is reduced or high zoom lenses are adopted to obtain higher resolution images, then detection precision is improved, but the field of view decreases and the workpiece cannot be captured completely
Solution Approach 1:
The imaging area is segmented into multiple sub-areas that are captured separately. Each sub-area can be imaged with sufficient resolution, and the combination of all sub-areas provides a complete view of the workpiece with high detection precision across the entire field of view
Solution Approach 2:
Instead of increasing resolution in one dimension (which reduces field of view), the system extends the field of view by capturing multiple sub-areas and merging them in the spatial domain. This dimensional approach allows both high resolution and large field of view to coexist
3Ease of manufacture
If conventional image composition process is used based on overlapping regions, then composite images can be formed, but it fails when images are non-overlapping or when characteristic information is not obvious enough
Solution Approach 1:
Known characteristic information is pre-marked on calibration devices at specific positions before imaging. This preliminary action enables the system to establish accurate correspondence relationships between multiple images even when they are non-overlapping, by using the pre-defined reference points for coordinate transformation
Solution Approach 2:
Calibration devices with known characteristic information serve as intermediaries between the image capturing devices and the workpiece. These intermediaries provide reference points that enable accurate coordinate transformation and image merging, especially when images do not overlap
Data Source
AI summary
The present disclosure provides an image merging method. The image merging method includes the following step. First, the calibration unit is provided, wherein a calibration device of the calibration unit includes a plurality of known characteristic information. The calibration device is captured. A conversion relationship is created. A relationship of positions of the images is analysis according to the conversion relationship. The images are formed. In additional, an image merging device is provided.