Pixel Classification for Depth Estimation Error Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current 3D digital imaging technologies face inaccuracies in depth estimation due to factors like rapid subject movement, semitransparent, specularly reflective, and retroreflective subjects, leading to unreliable downstream classification.
Innovation Solution
An upstream classification procedure labels segments of digital images affected by problem conditions, using a trained artificial neural network to mask or repair pixels before downstream classification, and computes repair values for corrupted depth or object classification pixels.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If depth estimation is performed on subjects with motion, semitransparent, specular, or retroreflective properties, then the imaging system can capture these subjects, but the depth estimation accuracy deteriorates
Solution Approach 1:
The system performs preliminary classification of pixels into trustworthy and untrustworthy categories before depth estimation. This upfront identification of problematic pixels (from semitransparent, specular, retroreflective, or motion-affected regions) allows the system to apply targeted corrections or exclusions during depth calculation, thereby maintaining accuracy across diverse subject types without sacrificing versatility
Solution Approach 2:
The patent applies different processing strategies to different regions of the image based on pixel classification. Trustworthy pixels undergo standard depth estimation, while untrustworthy pixels receive specialized handling such as exclusion from depth calculations or replacement with interpolated values from neighboring regions. This localized approach preserves overall system versatility while maintaining high accuracy in critical regions
2Measurement precision
If classification is performed on all pixels to identify corruption risks, then depth estimation accuracy improves, but processing time increases
Solution Approach 1:
The system performs classification on all pixels to ensure comprehensive identification of problematic regions, but then applies selective processing only to the subset of untrustworthy pixels. This partial action approach maintains high depth estimation accuracy by thoroughly identifying all potential issues, while reducing processing time by avoiding unnecessary complex operations on already-trustworthy pixels
3Reliability
If repair values are computed and applied to correct corrupted depth pixels, then depth image reliability improves, but system complexity increases
Solution Approach 1:
The system extracts and isolates the problematic pixels from the overall depth estimation process by identifying them through classification. These untrustworthy pixels are then handled separately through repair value computation or exclusion, removing their harmful influence on the final depth image. This extraction approach improves reliability by systematically addressing corruption sources while managing complexity through modular, targeted processing
Data Source
AI summary
A method to process a contributing digital image of a subject in an image-processing computer. The contributing digital image is received in a depth-resolving machine configured to furnish a depth image based at least in part on the contributing digital image. The contributing digital image is also received in a classification machine previously trained to classify a pixel of the contributing digital image as liable to corrupt a depth value of a corresponding pixel of the depth image. A repair value is computed for the depth value of the corresponding pixel of the depth image, which is then corrected based on the repair value and returned to the calling process.