Tread Pattern Image Classification using Convolutional Neural Network Based on Transfer Learning
Ying Liu, Shuai Zhang, Fuping Wang, Nam Ling · 2018
Tread pattern image classification helps in providing useful clues in police case solving and traffic accident management. To further boost-up the classification performance, this paper proposes a novel tread pattern classification algorithm using convolutional neural network (CNN) based on the idea of transfer learning. The algorithm consists of two parts: (1) Transfer the knowledge of a pre-trained CNN model on ImageNet dataset to produce a new model for the task of tread pattern classification, by fine-tuning the model parameters through back-propagation using tread pattern image data. The concept of transfer learning solves the problem of lacking large training dataset. (2) The features from multiple fully-connected layers are combined with different weights and used to train support vector machine (SVM) classifiers for image classification. Experimental results demonstrated the outstanding performance of the proposed algorithm over other existing methods for the task of tread pattern image classification.