Traffic light recognition is a critical requirement in autonomous driving. Incorrect recognition of traffic lights may lead to serious traffic accidents. Traditional traffic light recognition methods are categorized into model-based methods, semantic-segmentation-based methods, and object-detection-based methods. Model-based methods are strongly influenced by the lighting and environment. Although semantic-segmentation-based methods can detect traffic lights well, their time cost is high. Object-detection-based methods can satisfy the requirements of real-time traffic light recognition in terms of performance and time. This study proposes a real-time traffic light recognition algorithm that first extracts traffic lights from images using the YOLOv5s object detection algorithm based on the N_ResNet network and then classifies the state of the extracted traffic light images using the SVM classifier. To address the inconsistent size of traffic lights in real scenes and the large number of medium-size traffic lights, this study proposes a new network structure called N_ResNet. N_ResNet can effectively detect traffic lights of different sizes, especially those of medium size. It is similar to the three-layer ResNet network structure; the main difference is that the size of the convolutional kernel of the first convolutional layer is 3. We apply N_ResNet to the YOLOv5s object detection algorithm for traffic light detection. The results show that the proposed network achieves a significant improvement in the recall compared to DarkNet. The overall recall is improved by 2.66%, including 0.47% for traffic lights in the daytime and 11.32% for traffic lights in the nighttime. In addition, the accuracy of the SVM classifier reaches 92.2%.
Traditional hyperspectral image semantic segmentation algorithms can not fully utilize the spatial information or realize efficient segmentation with less sample data. In order to solve the above problems, a U-shaped hyperspectral semantic segmentation model (DCCaps-UNet) based on the depthwise separable and conditional convolution capsule network was proposed in this study. The whole network is an encoding–decoding structure. In the encoding part, image features are firstly fully extracted and fused. In the decoding part, images are then reconstructed by upsampling. In the encoding part, a dilated convolutional capsule block is proposed to fully acquire spatial information and deep features and reduce the calculation cost of dynamic routes using a conditional sliding window. A depthwise separable block is constructed to replace the common convolution layer in the traditional capsule network and efficiently reduce network parameters. After principal component analysis (PCA) dimension reduction and patch preprocessing, the proposed model was experimentally tested with Indian Pines and Pavia University public hyperspectral image datasets. The obtained segmentation results of various ground objects were analyzed and compared with those obtained with other semantic segmentation models. The proposed model performed better than other semantic segmentation methods and achieved higher segmentation accuracy with the same samples. Dice coefficients reached 0.9989 and 0.9999. The OA value can reach 99.92% and 100%, respectively, thus, verifying the effectiveness of the proposed model.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.