Symmetry features for license plate classification

CAAI Transactions on Intelligence Technology

article

Karpuravalli Srinivas Raghunandan¹  Palaiahnakote Shivakumara²  Lolika Padmanabhan³  Govindaraju Hemantha Kumar¹  Tong Lu⁴  Umapada Pal⁵ 

[1] Department of Studies in Computer Science, University of Mysore;Faculty of Computer Science and Information Technology, University of Malaya;PES Institute of Technology;National Key Lab for Novel Software Technology, Nanjing University;Computer Vision and Pattern Recognition Unit, Indian Statistical Institute

关键词: video signal processing;    edge detection;    image segmentation;    text analysis;    support vector machines;    optical character recognition;    image classification;    image colour analysis;    gradient methods;    image recognition;    feature extraction;    feature matrix;    symmetry features;    license plate classification;    high recognition rate;    statistical features;    global symmetry;    local symmetry;    global candidate stroke pixels;    common stroke pixels;    local candidate stroke pixels;    stroke pixel direction;    stroke width distance;    GVF opposite direction;    cursive text;    input license image;    cursive texts;    printed texts;    multitype images;    license plate images;    (B6135) Optical;    image and video signal processing;    (B6135E) Image recognition;    (C5260B) Computer vision and image processing techniques;    (C6170K) Knowledge engineering techniques;   

DOI  :  10.1049/trit.2018.1016

学科分类：数学（综合）

来源: Wiley

【 摘 要 】

Achieving high recognition rate for license plate images is challenging due to multi-type images. We present new symmetry features based on stroke width for classifying each input license image as private, taxi, cursive text, when they expand the symbols by writing and non-text such that an appropriate optical character recognition (OCR) can be chosen for enhancing recognition performance. The proposed method explores gradient vector flow (GVF) for defining symmetry features, namely, GVF opposite direction, stroke width distance, and stroke pixel direction. Stroke pixels in Canny and Sobel which satisfy the above symmetry features are called local candidate stroke pixels. Common stroke pixels of the local candidate stroke pixels are considered as the global candidate stroke pixels. Spatial distribution of stroke pixels in local and global symmetry are explored by generating a weighted proximity matrix to extract statistical features, namely, mean, standard deviation, median and standard deviation with respect the median. The feature matrix is finally fed to an support vector machine (SVM) classifier for classification. Experimental results on large datasets for classification show that the proposed method outperforms the existing methods. The usefulness and effectiveness of the proposed classification is demonstrated by conducting recognition experiments before and after classification.

【 授权许可】

CC BY|CC BY-ND|CC BY-NC|CC BY-NC-ND