Image Warping Interpolation for Missing Pixel Data Recovery
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Industrial scanners and barcode readers face limitations in determining surface features and dimensions due to missing pixel data caused by out-of-view regions, obstructions, or defective camera pixels, leading to reduced scanning efficiency and accuracy.
Innovation Solution
A method and system for image warping that utilizes a first and second imaging assembly to determine mapped pixel locations and values based on pixel neighborhood configurations, employing interpolation, extrapolation, or missing data pixel values to fill in missing data, and transform images to a common field of view.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If images are captured using imaging assemblies with limited field of view, then device complexity is reduced, but measurement precision deteriorates due to missing pixel data
Solution Approach 1:
The patent introduces an image processing system that acts as an intermediary between the captured images and the final measurement results. This system fills missing pixel data by interpolating from neighboring pixels and transforming multiple images with different missing data patterns into a composite image, thereby recovering measurement information without adding physical imaging hardware.
Solution Approach 2:
The patent creates a complete virtual copy of the object by combining multiple partial images. Each imaging assembly captures a portion of the object, and the processing system reconstructs the full object image by mapping and merging these partial views, effectively creating a complete digital representation without requiring a single comprehensive physical camera.
2Measurement precision
If multiple imaging assemblies are used to capture complete object data, then measurement precision improves, but device complexity increases
Solution Approach 1:
The patent merges multiple images captured by different imaging assemblies into a single composite image. The processing system aligns, transforms, and combines these images, filling in missing data from each view using data from other views, thereby achieving complete object coverage with a manageable number of cameras.
Solution Approach 2:
The patent transforms the problem from spatial coverage (requiring more cameras) to data dimensionality (combining multiple 2D images to reconstruct complete 3D object information). By using coordinate transformation and pixel mapping between different viewing angles, the system recovers missing information through mathematical transformation rather than additional hardware.
3Productivity
If images with missing data are combined directly, then productivity is improved by avoiding manual inspection, but measurement precision deteriorates due to compounded errors
Solution Approach 1:
The patent performs preliminary processing on each image before combining them. Missing pixels are identified and filled using interpolation from neighboring pixels in the same image, and coordinate transformations are pre-computed to align images properly. This preliminary preparation prevents error propagation when images are merged, maintaining measurement accuracy while enabling automated processing.
Data Source
AI summary
A method and system for performing image warping of images having missing pixel data. The method includes obtaining, by a first imaging assembly having a first field of view, a first image of one or more objects in the first field of view. The first image has one or more pixels with missing data pixel values. A processor determines a mapped pixel location for each pixel of the first image. The processor then identifies a pixel neighborhood having a pixel neighborhood configuration for each mapped pixel location. The processor further determines a mapped pixel value for each mapped pixel, each mapped pixel value based on the pixel neighborhood configuration. The warped image of mapped pixels and mapped pixel values is then presented to a user.