Image Collation Using Model Goodness-of-Fit Evaluation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image and three-dimensional data recognition systems face challenges in accurately collating images with varying environmental conditions due to instability in the model fitting process, leading to decreased recognition accuracy and increased risk of misidentification.
Innovation Solution
A collation apparatus and method that uses a variation model to estimate parameters for generating target and reference data, with a model goodness-of-fit evaluation to assess the accuracy of the fitting process, allowing for improved parameter estimation and reduced influence of variable elements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If model fitting process is used to estimate parameters for image variation, then recognition capability is improved, but stability of the fitting process deteriorates leading to decreased recognition accuracy
Solution Approach 1:
The patent introduces a model goodness-of-fit evaluation as an intermediary mechanism between the model fitting process and the recognition decision. This evaluation computes a goodness-of-fit value that mediates the unstable fitting results, providing a stable criterion for judgment. The goodness-of-fit evaluation acts as a buffer that transforms unstable parameter estimates into reliable recognition decisions.
Solution Approach 2:
The patent implements feedback by computing the model goodness-of-fit evaluation result and using it to determine whether to accept or reject the model fitting outcome. This feedback loop allows the system to adjust its recognition decision based on the quality of the fitting process, thereby stabilizing the overall recognition accuracy while maintaining adaptability.
2Measurement precision
If model fitting is performed to generate images under given conditions, then collation accuracy is improved, but computation time increases
Solution Approach 1:
The patent applies partial action by performing model fitting only when necessary for accurate collation, rather than always generating images under all possible conditions. The system selectively applies the computationally intensive model fitting process based on the specific recognition scenario, thereby maintaining high collation accuracy while reducing unnecessary computation time.
3Measurement precision
If variations of image luminance values caused by pose or illumination are reduced, then recognition accuracy is improved, but the ability to handle diverse environmental conditions deteriorates
Solution Approach 1:
The patent employs parameter changes by adjusting the model goodness-of-fit evaluation criteria based on the specific environmental conditions being analyzed. The system dynamically modifies evaluation parameters to account for variations in pose, illumination, and other environmental factors, thereby maintaining high recognition accuracy across diverse conditions without sacrificing adaptability.
Data Source
AI summary
A 2D model fitting means 1201 estimates values of parameters optimum to generate a probe model image similar to a probe image and a gallery model image similar to a gallery image with an image variation model 1204. At that time, among a plurality of parameters of the image variation model 1204, the value of a parameter of which sameness between the probe image and the gallery image is to be judged as a target parameter is set to be the same for both images. A model goodness-of-fit evaluation means 1202 computes a model goodness of fit for the probe model image and the gallery model image to the probe image and the gallery image under the estimated parameters. A collation means 1203 compares the model goodness of fit with a threshold value to judge the sameness between the probe image and the gallery image.