Guided Vehicle Exterior Damage Detection via Deep Learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In rental car and valet services, drivers face inconvenience and potential disputes due to the difficulty in capturing comprehensive images of vehicle exterior damage, especially when they are unfamiliar with areas prone to damage, leading to overlooked damages and lack of evidence for pre-existing issues.
Innovation Solution
A processor-implemented method using a mobile terminal with a camera to guide image capture of predetermined areas, combining RGB and depth images with a deep learning model to determine vehicle exterior damage, including areas with high damage likelihood or impact, and outputting results via an output device.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the driver captures the image while checking each portion of the exterior of the vehicle by himself/herself, then the driver can identify visible damage, but the process is inconvenient and time-consuming
Solution Approach 1:
The system enables the mobile terminal to automatically capture images of the vehicle exterior without requiring the driver to manually position and capture each area. The automatic image capture function performs the detection task independently, eliminating the inconvenience of manual operation while maintaining comprehensive damage detection capability
Solution Approach 2:
The patent replaces the manual mechanical process of the driver physically inspecting and capturing images with an automated system using deep learning models and computer vision algorithms. The system automatically processes images to detect damage, substituting human manual inspection with intelligent automated analysis
2Productivity
If the driver captures images manually without knowing damage-prone areas, then the process can be simple, but a lot of time is required and damage areas may be missed
Solution Approach 1:
The system pre-stores baseline images of the vehicle exterior taken at the beginning of the rental period. These preliminary images serve as a reference for automatic comparison, enabling the system to efficiently identify any changes or damage without requiring the driver to manually inspect each area during handover
Solution Approach 2:
The deep learning model automatically compares the current vehicle exterior images with the pre-stored baseline images, providing feedback on detected differences and potential damage. This automated feedback mechanism ensures comprehensive damage detection while reducing the time required for manual inspection
3Ease of operation
If the driver relies on naked eye inspection, then the process is simple, but damaged portions difficult to identify are overlooked
Solution Approach 1:
The system creates digital copies of the vehicle exterior through captured images and uses deep learning models to analyze these copies. This allows detailed examination of fine damages such as scratches and dents without requiring the driver to physically inspect each area, maintaining operational simplicity while enhancing detection precision
Solution Approach 2:
The patent replaces the human naked eye inspection mechanism with an automated computer vision system using deep learning algorithms. This substitution enables the detection of fine damages that are difficult to identify manually, such as small scratches and dents, while keeping the process simple for the user
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This method streamlines the process of identifying vehicle exterior damage, reducing the likelihood of disputes by providing accurate and comprehensive damage assessments, including types such as scratches, dents, and cracks, through guided image capture and machine learning analysis.
Implementation Method 1
a camera mounted on a mobile terminal
Implementation Method 2
The camera may include a time of flight (ToF) camera configured to capture the depth image
Data Source
AI summary
A processor implemented method including outputting guide information to guide a capture of an image of a predetermined area via a camera of a mobile terminal including the processor, inputting a first exterior image of a vehicle, the first exterior image being captured based on the output guide information and a second exterior image of the vehicle stored in advance to a processor including a deep learning model, matching the first exterior image and the second exterior image with each other to acquire a matched image, masking a detected area from the predetermined image within the matched image as a masked area, and determining whether an exterior of the vehicle has been damaged and a type of damage based on the masked area.