Image-Based Vehicle Classification Using Neural Network Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for measuring traffic flow and controlling traffic lights are inefficient and labor-intensive, relying on manual vehicle counting on roads.
Innovation Solution
An image-based vehicle classification system utilizing a camera and image server, which captures and processes images to segment, thin, erode, and dilate vehicle images, and employs a neural network classifier to automatically classify vehicles into large-size cars, passenger cars, and motorcycles, thereby facilitating automatic traffic management.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If manual vehicle counting is used to measure traffic flow, then the method is simple to implement, but it is inefficient and labor-intensive
Solution Approach 1:
The patent replaces manual mechanical counting with an automated image processing system that uses computer vision algorithms to detect, segment, and classify vehicles from camera images, eliminating the need for human operators to manually count vehicles
Solution Approach 2:
The system enables self-service automation where the image processing system automatically performs vehicle detection, classification, and counting without human intervention, with the processor independently executing the entire workflow from image acquisition to traffic flow measurement
2Measurement precision
If image processing with multiple steps (segmentation, thinning, erosion, dilation) is performed, then vehicle classification accuracy is improved, but processing complexity increases
Solution Approach 1:
The patent applies image segmentation to divide the image into foreground (vehicle) and background portions, then further segments overlapping vehicles into separate image parts through thinning and erosion operations, enabling accurate individual vehicle classification
Solution Approach 2:
The patent performs preliminary image processing steps (segmentation, thinning, erosion, dilation) before classification to prepare the image data, ensuring that vehicle image parts are properly separated and shaped before being fed into the neural network classifier
3Adaptability or versatility
If a neural network classifier is used for vehicle classification, then classification capability is enhanced, but computational requirements and processing time increase
Solution Approach 1:
The patent performs preliminary image processing (segmentation, thinning, erosion, dilation) to prepare clean, separated vehicle image parts before classification, reducing the computational burden on the neural network and enabling faster processing
Solution Approach 2:
By segmenting overlapping vehicles into separate image parts through morphological operations, the system enables the neural network to classify each vehicle independently and efficiently, rather than attempting to classify complex overlapping regions as a single unit
Data Source
AI summary
An image-based vehicle classification system includes a camera and an image server connected to the camera. The camera captures images of a road to result in an image stream. The image server includes a processor for receiving the image stream from the camera. For each of the images of the image stream, the processor performs image segmentation, a thinning process, an erosion process and a dilation process, and classifies, by a neural network classifier, a vehicle image part contained in the image into one of a large-size car class, a passenger car class and a motorcycle class when it is determined that the vehicle image is crossing an imaginary line set in advance in the image for counting vehicles.