Odometer Mileage Extraction From Variable-Quality Vehicle Images
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing OCR systems struggle to accurately extract odometer mileage from images due to variations in color, intensity, and texture, and require additional hardware or human verification, leading to inefficiencies and increased costs.
Innovation Solution
A system utilizing a trained Faster RCNN model for object detection and post-processing to identify and extract odometer characters from images, leveraging image augmentation and character recognition to improve accuracy and reduce reliance on additional sensors or human verification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional OCR systems are used to extract odometer mileage, then the system can process images, but the accuracy deteriorates due to variations in color, intensity, and texture
Solution Approach 1:
The system applies image augmentation techniques that transform images by adjusting parameters such as color, intensity, and texture. This creates varied training data that teaches the model to recognize odometer readings despite variations in lighting, color, and display characteristics, thereby improving both accuracy and reliability
Solution Approach 2:
The system creates synthetic copies of odometer images through image augmentation, generating training data that replicates various real-world conditions. These synthetic images allow the model to learn patterns from multiple variations without requiring physical manipulation of the actual odometers
2Reliability
If additional hardware or human verification is used to improve accuracy, then the reliability improves, but the device complexity and cost increase
Solution Approach 1:
The system enables self-service extraction by using standard image capturing devices (smartphones, cameras) that users already possess. The machine learning model automatically processes the images without requiring additional specialized hardware or human verification, making the system both reliable and simple to use
Solution Approach 2:
The system replaces mechanical or human-based verification methods with an automated machine learning model. Instead of requiring physical inspection or human review, the neural network performs the extraction and verification automatically, reducing complexity while maintaining or improving reliability
3Measurement precision
If manual verification processes are implemented, then the accuracy improves, but the productivity deteriorates due to time consumption
Solution Approach 1:
The system performs continuous automated extraction as images are captured, eliminating interruptions for manual verification. The machine learning model processes images in real-time or near-real-time, maintaining continuous workflow and significantly improving productivity while preserving accuracy through consistent automated processing
Data Source
AI summary
Systems and methods are provided for determining information, such as characters, from an image of a vehicle odometer using a trained character recognition object detection model. Methods and systems are included for training a character recognition object detection model to identify characters in the image of a vehicle odometer by repeatedly receiving an image, augmenting the image, identifying odometer character regions and characters, and comparing the identified odometer character regions and characters with those on annotated training images, and updating the model. Cloud-based and mobile application systems are provided for receiving an image from a user, using the trained character recognition object detection model to output odometer character regions and, for each character region, a class label and a probability, and using a post-processing application to determine the odometer mileage number based on the output of the trained character recognition object detection model.