Vehicle Image Comparison for Defect Detection With Fewer False Alarms
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing vehicle image analysis methods are cumbersome, time-consuming, and computationally inefficient, particularly in identifying real physical changes (DOI) while excluding false alarms (FA) due to varying imaging conditions.
Innovation Solution
A computerized method and system for vehicle image comparison using deep learning models for segmentation, reference image selection, and similarity analysis to identify DOI while minimizing FA, employing segmentation DNN, encoding DNN, similarity DNN, and comparison DNN to process and compare vehicle images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If manual inspection methods are used for vehicle images, then flexibility and adaptability are maintained, but inspection efficiency is low and time-consuming
Solution Approach 1:
The patent creates a standardized digital copy of the inspection process by capturing vehicle images at predetermined locations and comparing them with reference images. This copying approach allows automated inspection while maintaining the flexibility of manual methods through standardized data reproduction rather than physical replication.
Solution Approach 2:
The patent replaces manual mechanical inspection with an automated image processing system that uses cameras, processors, and comparison algorithms. The mechanical inspection process is substituted with electronic image capture, digital storage, and automated comparison to determine defect presence.
2Measurement precision
If automated image comparison systems are implemented, then inspection speed increases, but accuracy may be compromised due to inability to handle complex defect patterns
Solution Approach 1:
The patent segments the vehicle inspection process into multiple predetermined locations and standardized comparison points. By dividing the complex inspection task into discrete, standardized segments at specific locations, the system can efficiently compare images while maintaining accuracy through systematic coverage of all critical areas.
Solution Approach 2:
The patent changes the parameters of image comparison by establishing predetermined comparison locations and reference images with standardized characteristics. This parameter standardization allows automated systems to accurately detect defects by comparing images against known reference parameters at specific locations.
3Reliability
If standardized comparison locations are established, then inspection consistency improves, but adaptability to different vehicle types decreases
Solution Approach 1:
The patent creates a universal inspection system that can accommodate different vehicle types by establishing standardized comparison locations that can be applied across multiple vehicle models. The same fundamental comparison methodology works for different vehicles, with specific locations and reference images adjusted as needed to maintain both consistency and adaptability.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
There are provided a system and method of vehicle image comparison, the method including: obtaining an input image comprising a plurality of image portions; retrieving a set of reference images; for each image portion, searching for a best matching reference portion in the set of reference images, comprising: i) for each given reference image: identifying a reference region; using a similarity model on the given image portion and the reference region to obtain a similarity map indicating a similarity between the image portion and a respective reference image portion; and selecting a reference image portion with the best similarity as a reference portion candidate; and ii) selecting the best matching reference portion; and comparing each given image portion with the best matching reference portion using a comparison model, giving rise to a difference map indicating probability of presence of DOI in the given image portion.