Search

Counting of passengers entering and exiting means of transport is one of the basic functionalities of passenger flow monitoring systems. Exact numbers of passengers are important in areas such as public transport surveillance, passenger flow prediction, transport planning, and transport vehicle load monitoring. To allow mass utilization of passenger flow monitoring systems, their cost must be low. As the overall price is mainly given by prices of the used sensor and processing unit, we propose the utilization of a visible spectrum camera and data processing algorithms of low time complexity to ensure a low price of the final product. To guarantee the anonymity of passengers, we suggest orthogonal scanning of a scene. As the precision of the counting is relevantly influenced by the precision of passenger recognition, we focus on the development of an appropriate recognition method. We present two opposite approaches which can be used for the passenger recognition in means of transport with and without entrance steps, or with split level flooring. The first approach is the utilization of an appropriate convolutional neural network (ConvNet), which is currently the prevailing approach in computer vision. The second approach is the utilization of histograms of oriented gradients (HOG) features in combination with a support vector machine classifier. This approach is a representative of classical methods. We study both approaches in terms of practical applications, where real-time processing of data is one of the basic assumptions. Specifically, we examine classification performance and time complexity of the approaches for various topologies and settings, respectively. For this purpose, we form and make publicly available a large-scale, class-balanced dataset of labelled RGB images. We demonstrate that, compared to ConvNets, the HOG-based passenger recognition is more suitable for practical applications. For an appropriate setting, it defeats the ConvNets in terms of time complexity while keeping excellent classification performance. To allow verification of theoretical findings, we construct an engineering prototype of the system., Author(s): Pavel Skrabanek (corresponding author) [1]; Petr Dolezel [2]; Zdenek Nemec [2]; Dominik Stursa [2] 1. Introduction In passenger transport, person flow monitoring has an indispensable importance. In some areas [...]

The article presents a brief overview of new electric-hydraulic systems of three selected earthmoving machines developed by means of mathematical modeling and simulation. These machines were as follows: a road tire roller, an excavator and a wheel loader. Their common feature was that a special innovative system was developed for each of them by Bosch Rexroth company (Brno) in cooperation with BUT (Brno University of Technology) using mathematical modeling of hydraulic, mechanical and electrical processes in the machines. Each of these machines represents an innovation: a module for the recovery of kinetic ("braking") energy was developed for the road roller, and an emission-free electric battery drive for the excavator and the loader as a replacement for the diesel engine drive. [ABSTRACT FROM AUTHOR]

This letter presents the design of an overmoded $X$ -band RF window for use in high-power microwave tubes, such as klystron, magnetron, and linear accelerator (LINAC). The cylindrical section of the window containing dielectric barrier operates in the TE022 mode. Conversion from TE10 mode to TE022 mode is achieved using four rectangular slots made on the narrow walls of the rectangular waveguide sections. The simulation results for the proposed design in CST and HFSS are in good agreement. The return loss, insertion loss, and bandwidth are obtained as 52.76 dB, 0.06 dB, and 41.4 MHz, respectively, at 9.3 GHz. Coupled electromagnetic, thermal, and structural simulations are performed to study the multiphysics performance of the RF window. The multipactor analysis is carried out in SPARK3D to determine the multipactor threshold. The proposed TE022 mode RF window can operate safely at a 20.75 MW of peak power. [ABSTRACT FROM AUTHOR]

This paper briefly describes some features of control of the hydrostatic recovery module developed within the framework of a research project. This project was focused on research and development of a new product - a hydrostatic system for energy recovery through breaking and start-up of commercial vehicles using control for energy transfer and optimization based on mathematical modeling of the system activity. A contribution of described solution of control system of hydraulic recovery module of the road roller can be seen in the fact that this solution enabled fuel savings during recovery up to 26.6% at low operation speed of 9 km/h. [ABSTRACT FROM AUTHOR]

The papers deals with a modeling of a Time- Difference of Arrival system for a subscriber station localization, based on the 802.11 standard wireless network. In the case of severe multipath effects the standard TDOA estimation methods, based on correlation of signals, received by conveniently displaced receiving stations show large errors. Thus, a new algorithm is proposed using received signals decomposition to a set of delayed replicas. This represents a linear estimation of reflected signals amplitudes. The described method leads to a better estimation of time differences of the signals, propagating on the direct paths between the emitter and the receiving stations. [ABSTRACT FROM AUTHOR]

In the field of computer vision, object detection consists of automatically finding objects in images by giving their positions. The most common fields of application are safety systems (pedestrian detection, identification of behavior) and control systems. Another important application is head/person detection, which is the primary material for road safety, rescue, surveillance, etc. In this study, we developed a new approach based on two parallel Deeplapv3+ to improve the performance of the person detection system. For the implementation of our semantic segmentation model, a working methodology with two types of ground truths extracted from the bounding boxes given by the original ground truths was established. The approach has been implemented in our two private datasets as well as in a public dataset. To show the performance of the proposed system, a comparative analysis was carried out on two deep learning semantic segmentation state-of-art models: SegNet and U-Net. By achieving 99.14% of global accuracy, the result demonstrated that the developed strategy could be an efficient way to build a deep neural network model for semantic segmentation. This strategy can be used, not only for the detection of the human head but also be applied in several semantic segmentation applications. [ABSTRACT FROM AUTHOR]