License Plate Localization via Edge Detection and Dual-Binarization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current automatic license plate recognition (ALPR) systems are inefficient in localizing license plates across different demographics and languages due to variations in script, font, and color, and often require high computational resources, making them unsuitable for real-time applications with varying background and foreground colors.
Innovation Solution
A system and method for language-independent automatic real-time license plate localization using an imaging processor that applies edge region detection, dual-binarization, and grouping units to identify and isolate license plates, employing down-sampling, motion-based search-space reduction, and morphological operations to achieve computational efficiency without sacrificing accuracy, and operates independently of illumination conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If image thresholding/binarization techniques are used for license plate localization, then the system can process images efficiently, but it fails when license plates have different foreground-background color combinations and multi-colored textures
Solution Approach 1:
The system transforms the input image from RGB color space to HSV (Hue, Saturation, Value) color space, changing the parameter representation of colors. This allows the system to work with hue information independently of brightness and saturation variations, enabling the license plate detection to be invariant to different color combinations and lighting conditions while maintaining processing efficiency
Solution Approach 2:
The license plate detection process is divided into multiple independent stages: edge detection to identify plate boundaries, contour analysis to determine plate shape and orientation, and character recognition to read the license number. This segmentation allows each stage to be optimized independently and makes the system adaptable to various color schemes without requiring complete redesign
2Measurement precision
If template-matching with overlapping sliding-window protocol is used for license plate detection, then the system can effectively identify known license plate types, but it results in high computational cost and reduced processing speed
Solution Approach 1:
The system performs preliminary edge detection and contour analysis to identify the license plate region and its orientation before attempting character recognition. This preliminary action narrows down the search space and eliminates the need for exhaustive template matching across the entire image, significantly reducing computational cost while maintaining detection accuracy
Solution Approach 2:
Instead of applying full template matching to the entire image, the system applies template matching only to the identified license plate region after edge and contour analysis. This partial action approach maintains high detection accuracy for known plate types while dramatically reducing the computational burden by limiting the search space
3Measurement precision
If deep learning technology is used for license plate detection, then the system achieves better accuracy, but it requires high computing environment which is not suitable for real-time applications
Solution Approach 1:
The system replaces complex deep learning mechanical systems with a sequence of simpler image processing operations: edge detection using Sobel operators, contour analysis using geometric transformations, and template matching. This substitution achieves comparable detection accuracy while dramatically reducing computational requirements and enabling real-time processing on standard hardware
4Productivity
If vertical and horizontal edge projection based methods are used for license plate localization, then the system can process images in near-real time, but it fails when the image region contains too much edge-like structures around the license plate location
Solution Approach 1:
The system transitions from 2D edge projection analysis to 3D surface analysis by computing the normal vectors of detected edges and analyzing their spatial distribution in three-dimensional space. This dimensional change allows the system to distinguish between edges that belong to the license plate and edges that are part of the background, improving detection reliability in complex scenes while maintaining near-real-time processing speed
Data Source
AI summary
The present invention discloses a system and a method for language independent automatic license plate localization by analysis of plurality of images in real-time under day-light condition without using any external light. In one embodiment, the system can work without any spatiality constraints and/or demographic considerations without any restriction on jurisdiction and can effectively localize license plates (LPs) of any type consisting of alpha-numeric characters and symbols. In other embodiment, methods for search-space reduction system based on motion-based filtration and LPs localization based on edge-active-region filtration with high frame-per-second (FPS) throughput are described. In another embodiment, a dual-binarization scheme is described for color invariant LP localization.