Traffic sign recognition by combining global and local features based on semi‐supervised classification

Abstract

The legibility of traffic signs has been considered from the beginning of design, and traffic signs are easy to identify for humans. For computer systems, however, identifying traffic signs still poses a challenging problem. Both image-processing and machine-learning algorithms are constantly improving, aimed at better solving this problem. However, with a dramatic increase in the number of traffic signs, labelling a large amount of training data means high cost. Therefore, how to use a small number of labelled traffic sign data reasonably to build an efficient and high-quality traffic sign recognition (TSR) model in the Internet-of-things-based (IOT-based) transport system has been an urgent research goal. Here, the authors propose a novel semi-supervised learning approach combining global and local features for TSR in an IOT-based transport system. In their approach, histograms of oriented gradient, colour histograms (CH), and edge features (EF) are used to build different feature spaces. Meanwhile, on the unlabelled samples, a fusion feature space is found to alleviate the differences between different feature spaces. Extensive evaluations on a collection of signs from the German Traffic Sign Recognition Benchmark (GTSRB) dataset shows that the proposed approach outperforms the others and provides a potential solution for practical applications.

Robust Semi-Supervised Traffic Sign Recognition via Self-Training and Weakly-Supervised Learning

“…To compare the proposed approach with some state-of-the-art semi-supervised algorithms that were evaluated on the GTSRB including TSCA co-training [1] and multiple feature representation [46], TSCA co-training [1] proposed by Hillebrand et al was trained on just 14 classes. A recognition accuracy of 99.71% was obtained for precision, recall, and F1-score for the GTSRB and for the BTSC, 98.95%, 98.87%, 98.86%, respectively.…”

“…When it comes to the application of semi-supervised learning methods for the traffic sign recognition task, a few literature works can be found. He et al proposed a novel semi-supervised learning method that combined global and local features for traffic sign recognition in an Internet of Things-based transport system [46]. In that research, different feature spaces were built utilizing approaches such as the histogram of oriented gradients (HOG), color histograms (CH), and edge features (EF) for the labeled aspect and for the unlabeled data samples.…”

“…Hillebrand et al proposed applying semi-supervised co-training to classify German traffic signs [1] prior to the study by He et al In that research, Hillebrand et al deployed an iterative co-training process where the most informative samples from a given pool of unlabeled traffic signs were automatically selected and then classified by two classifiers, which generated labels for each other [1]. Extensive experiments were conducted on 14 classes of German traffic signs to obtain an accuracy of 98.0%, which would later be improved by the work of He et al [46].…”

Semi-Supervised Learning-Based Image Denoising for Big Data

“…To make the sample distribution after dimensionality reduction and in high-dimensional space can be close enough to each other, the p dispersion is chosen as the loss function of SNE in the optimization process, as shown in formula (8).…”

“…Therefore, it is necessary to reduce the noise with the help of scientific algorithms. For the human visual nerves, the de-noised images can show clearer and more accurate intent, which helps visual recognition and cognition [8]. Because of various internal and external factors, it is difficult to find a noise-free image that has varying degrees of contamination, which makes it difficult to distinguish relevant details [9].…”