Projection Image Distortion Correction via Vertex Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing projection image correction techniques fail to accurately correct distortions, particularly when the projection image exceeds the screen dimensions, as they rely on reducing image dimensions which can lead to incomplete vertex detection and inadequate correction.
Innovation Solution
An image processing apparatus that detects vertices in a captured projection image, reduces image size if necessary, and determines a distortion correction parameter based on successful vertex detection in a size-reduced image, allowing for precise correction of projection image distortions by calculating vertex positions and applying a distortion correction parameter.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the projection image dimensions are reduced to enable vertex detection, then vertex detection becomes possible, but the correction accuracy deteriorates due to loss of full-size image information
Solution Approach 1:
The patent applies preliminary action by first reducing the projection image dimensions to a state where vertex detection becomes possible, then performing vertex detection on the reduced image, and finally using those detection results to calculate distortion correction parameters for the full-size image. This sequential approach allows vertex detection to be performed on a manageable scale while still enabling correction of the original full-size projection image.
2Loss of information
If the projection image is captured at full size, then complete image information is obtained, but vertex detection becomes impossible when the image exceeds screen dimensions
Solution Approach 1:
The patent applies segmentation by dividing the correction process into distinct stages: first capturing the projection image at reduced dimensions where vertex detection is feasible, then using those detected vertices to calculate correction parameters that will be applied to the full-size image. This segmentation allows vertex detection to be performed on a manageable portion of the image while still enabling complete image correction.
3Reliability
If multiple reduced-size projection images are captured at different dimensions, then vertex detection success rate increases, but the processing time and complexity increase
Solution Approach 1:
The patent applies dynamics by implementing an adaptive process that captures projection images at multiple reduced dimensions and dynamically selects the most appropriate image for vertex detection based on detection success. The system adjusts the capture dimensions and selection criteria during the correction process to optimize both detection reliability and processing efficiency, rather than using a fixed predetermined approach.
Data Source
AI summary
An image processing apparatus includes a detection unit, a size changing unit, and a determining unit. The detection unit detects, from a first captured image obtained by capturing a projection image projected by a projection unit, a vertex of the projection image. The size changing unit reduces a size of the projection image in response to the detection unit not detecting a predetermined number of vertexes of the projection image from the first captured image. The determining unit determines a distortion correction parameter to be used for correcting a distortion of a projection image projected by the projection unit, in accordance with a detection result of vertexes by the detection unit for a second captured image obtained by capturing the size-reduced projection image in a case where the detection unit detects the predetermined number of vertexes of the size-reduced projection image from the second captured image.