Intelligent transportation systems (ITS) urgently need to realize vehicle identification, dynamic monitoring, and traffic flow monitoring under high-speed motion conditions. Vehicle tracking based on radio frequency identification (RFID) and electronic vehicle identification (EVI) can obtain continuous observation data for a long period of time, and the acquisition accuracy is relatively high, which is conducive to the discovery of rules. The data can provide key information for urban traffic decision-making research. In this paper, an RFID tag motion trajectory tracking method based on RF multiple features for ITS is proposed to analyze the movement trajectory of vehicles at important checkpoints. The method analyzes the accurate relationship between the RSSI, phase differences, and driving distances of the tag. It utilizes the information weight method to obtain the weights of multiple RF characteristics at different distances. Then, it calculates the center point of the common area where the vehicle may move under multi-antenna conditions, confirming the actual position of the vehicle. The experimental results show that the average positioning error of moving RFID tags based on dual-frequency signal phase differences and RSSI is less than 17 cm. This method can provide real-time, high-precision vehicle positioning and trajectory tracking solutions for ITS application scenarios such as parking guidance, unmanned vehicle route monitoring, and vehicle lane change detection.

The lack of a testing framework for various indoor positioning technologies brings huge challenges to the systematic and fair evaluation of positioning systems, which greatly hinders the development and industrialization of indoor positioning technology. In order to solve this problem, this article refers to international standards, such as ISO/IEC 18305, and uses the China Electronics Standardization Institute’s rich experience in indoor positioning technology research and testing to build a universal positioning performance testing and evaluation framework. First, this paper introduces the experimental environment in detail from the aspects of the coordinate system definition, test point selection, building type definition, motion mode definition, and motion trajectory setting. Then, this paper comprehensively measures performance evaluation indicators from dimensions such as the accuracy index, relative accuracy, startup time, fault tolerance, power consumption, size, and cost. Finally, this paper elaborates on the testing methods and processes of positioning precision, accuracy, relative accuracy, floor identification, indoor–outdoor distinction, latency, relative accuracy, success rate, and movement speed tests.

In recent years, deep convolutional neural networks (DCNNs) have made strong progress on single-image super-resolution (SISR). Most of them work in a fully supervised way: they learn a non-linear mapping from a low-resolution (LR) image to its high-resolution (HR) version. The main problem is that this needs ground-truth HR images for supervision, which makes these methods hard to use widely. To ease this problem, we propose a weakly supervised framework named AASPNet (Adaptive Attention-Based Symmetric Progressive Network). It has two parts: an upsampling process and a downsampling process. Each part is built from several training units, but only the first unit needs supervision. In other words, with only $\times 2$ supervision, AASPNet can produce HR results at $\times 4$ and $\times 8$. The symmetry between the two parts is used in two ways. First, it gives consistency constraints that help recover more accurate HR images. Second, a cross self-supervision recovers HR images in a weakly supervised way, using both image and feature consistency between the upsampling and downsampling parts. We also add an adaptive attention strategy to pick out useful features, so the HR images get better. Experiments on several benchmark datasets show that our weakly supervised AASPNet reaches results that are close to state-of-the-art fully supervised DCNN methods.