Image Matching Robustness Against Illumination Shading
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Image matching techniques are susceptible to illumination variations, leading to inaccurate matching, especially for objects with flat regions and few characteristic patterns, and require numerous reference images for improved accuracy, which is time-consuming and difficult to set up.
Innovation Solution
A data processing device computes difference values between pixels, extracts pixel groups influenced by shading, computes representative values statistically, and determines feature values to reduce the influence of shading, enabling robust image matching under varying illumination.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If increment sign correlation is used for image matching, then robustness against illumination variation is improved, but accuracy deteriorates due to shading influence from object shape
Solution Approach 1:
The patent introduces an intermediary process between raw image data and matching comparison: it computes difference values between adjacent pixels, then uses statistical processing (median filtering) to create a shaded image that removes shading artifacts. This intermediary representation allows the system to maintain robustness against illumination variation while eliminating the harmful shading effect that degrades accuracy.
Solution Approach 2:
The patent extracts only the essential information needed for accurate matching by computing difference values between adjacent pixels and then applying statistical processing to remove shading. This extraction process separates the useful structural information from the harmful illumination-induced shading, allowing accurate matching without being influenced by illumination variations or object shape-induced shading.
2Measurement precision
If multiple reference images are used to improve matching accuracy, then precision is improved, but device complexity and setup difficulty increase
Solution Approach 1:
The patent changes the parameter representation from using multiple reference images to using a single reference image processed through statistical operations. By computing difference values and applying median filtering to create a shaded image, the system achieves the accuracy benefits of multiple references while maintaining the simplicity of a single-reference setup.
Solution Approach 2:
The system performs self-service by automatically removing shading artifacts through statistical processing of the reference image itself. The median filtering operation uses the data within the reference image to eliminate shading, making the system robust without requiring external multiple reference images or complex setup procedures.
3Measurement precision
If multiple reference images are collected to improve matching accuracy, then precision is improved, but loss of time and setup effort increase
Solution Approach 1:
The patent applies preliminary action by pre-processing the reference image through statistical operations (difference value computation and median filtering) to create a shaded image that is robust against illumination variations. This preliminary processing eliminates the need for time-consuming collection of multiple reference images under different lighting conditions, as a single pre-processed reference image suffices.
Data Source
AI summary
A data processing device 300 according to the present invention comprises difference value computing means 402, 412 that computes a difference value between a pixel value of a target pixel that is each pixel contained in an image and a pixel value of a pixel that is present at a predetermined neighboring relative position of the target pixel, representative value computing means 403, 413 that extracts a pixel group containing pixels that are similarly influenced by shading due to light from the image with respect to each pixel of the image and computes a representative value of difference values of the pixel group according to a statistical technique, feature value computing means 404, 414 that computes the feature value with respect to each pixel contained in the image based on comparison between the difference value with respect to each pixel and the representative value of difference values of the pixel group, and similarity determining means 301 that determines a similarity between the image and a predetermined image based on the feature value with respect to each pixel extracted by the feature value computing means.