Recommendations for evaluating the performance of background subtraction algorithms for surveillance systems
- PDF / 2,319,288 Bytes
- 34 Pages / 439.642 x 666.49 pts Page_size
- 112 Downloads / 174 Views
Recommendations for evaluating the performance of background subtraction algorithms for surveillance systems Silvio Ricardo Rodrigues Sanches1 · Antonio Carlos Sementille2 · Ivan Abdo Aguilar3 · Valdinei Freire4 Received: 8 March 2020 / Revised: 26 July 2020 / Accepted: 9 September 2020 / © Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract Background subtraction is a prerequisite for a wide range of applications, including video surveillance systems. A significant number of algorithms are often developed and published in different publication mediums in the area, such as workshops, symposiums, conferences, and journals. An important task in presenting a new background subtraction algorithms is to clearly show that its performance outperforms the performance of the state-of-the-art algorithms. In this paper, we present recommendations on how to evaluate the performance of background subtraction algorithms for surveillance systems. We identified, through a systematic mapping, the key steps and components of this evaluation process – procedures, methods, and tools – most used by the authors in each of these steps. Considering this statistical analysis, we perform a theoretical analysis of the most used key components to identify their pros and cons. Then, we define a set of recommendations that aim to standardize and clarify the performance evaluation process of a new background subtraction algorithm. Keywords Background subtraction · Performance assessment · Recommendations · Surveillance systems Silvio Ricardo Rodrigues Sanches
[email protected] Antonio Carlos Sementille [email protected] Ivan Abdo Aguilar [email protected] Valdinei Freire [email protected] 1
Universidade Tecnol´ogica Federal do Paran´a, Corn´elio Proc´opio, Brazil
2
Universidade Estadual Paulista “J´ulio de Mesquita Filho”, Bauru, Brazil
3
Simon Fraser University, Burnaby, Canada
4
Universidade de S˜ao Paulo, S˜ao Paulo, Brazil
Multimedia Tools and Applications
1 Introduction Image segmentation is the task of subdividing an image into its constituent regions (or objects) [41]. Some of these regions are formed by pixels that belong to elements that are of interest to a specific application. An element of interest can be, for example, a person in applications such as people tracking [135], people counting [73] or fall detection [59]. Therefore, the regions of the image that contain pixels that belong to people are the ones that matter for the application. On the other hand, in applications such as vehicle counting [126] or accident detection [85], vehicles are the elements of interest. In these applications, segmentation is the first step of a process that uses computational intelligence algorithms in the later steps. In addition to the applications mentioned, the segmentation of images or video frames is a prerequisite for applications in the various domains [3, 29, 37, 97]. Surveillance systems, for example, need real-time segmentation to extract all the moving objects in the frames of a vi
Data Loading...
