Person Detection for an Orthogonally Placed Monocular Camera

Abstract

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.

CSU-Net: Contour Semantic Segmentation Self-Enhancement for Human Head Detection

“…This work is based on a long-term collaboration between the authors' team and industry partners. The achieved results follow the previous publication [9], where the problem of classification and detection of persons in visual data was solved using HOG descriptors; and the publication [10], where an improved DeepLabv3+ semantic segmentation approach was proposed to detect a human head in an RGB image.…”

New End-to-End Strategy Based on DeepLabv3+ Semantic Segmentation for Human Head Detection

“…Skrabanek et al [31] proposed a head detection system through a pipeline of vision passenger recognition systems based on ConvNets (using five different architectures) to ensure both feature extraction and classification. In addition, Khan et al [32] presented a head detection deep model-based method.…”

Multiple Objects Localization Using Image Segmentation with U-Net

“…with CNNs, we propose an approach based on creation of a schematic image leading to image segmentation. This basis for further use in different applications has been established in previous authors' publications [19], [20].…”