Image Processing and Post-Data Mining Processing for Security in Industrial Applications: Security in Industry

Image Processing and Post-Data Mining Processing for Security in Industrial Applications: Security in Industry

Alessandro Massaro (Dyrecta Lab srl, Italy) and Angelo Galiano (Dyrecta Lab srl, Italy)
DOI: 10.4018/978-1-7998-1290-6.ch006
OnDemand PDF Download:
No Current Special Offers


The chapter analyzes scientific approaches suitable for industrial security involving environment health monitoring and safety production control. In particular, it discusses data mining algorithms able to add hidden information important for security improvement. In particular k-means and artificial intelligence algorithms are applied in different cases of study by discussing the procedures useful to set the model including image processing and post clustering processing facilities. The chapter is focused on the discussion of information provided by data mining results. The proposed model is matched with different architectures involving different industrial applications such as biometric classification and transport security, railway inspections, video surveillance image processing, and quarry risk evaluation. These architectures refer to specific industry projects. As advanced applications, a clustering analysis approach applied on thermal radiometric images and a dynamic contour extraction process suitable for oil spill monitoring are proposed.
Chapter Preview

Introduction: Research Topics And Applications

Image and data processing in security systems are important research topics in different application fields. In order to improve security detection it is important to preliminary study the best requirements referring to a specific case of study. According to this, in this chapter are discussed some cases of industry research projects, by highlighting the data processing algorithms and security procedures. In this framework is analysed the state of the art suggesting the research topics of the proposed projects. Specifically, some works in literature analyses face recognition systems (Massaro et al. 2016) by creating a training dataset by means of synthetic data generated by original ones thus increasing the efficiency of the face recognition algorithm. Face recognition can be applied in different security applications involving different hardware and technologies, including mobile devices (Galiano et al. 2017). Another interesting research topic is the video surveillance processing and optimization, where a critical aspect is the edge detection of a body in motion which can be affected by different aspects such as hardware sensitivity and environment noise (Massaro, Vitti et al. 2018). In transport sector security and monitoring can be improved by (i) predictive maintenance, (ii) by data mining algorithms supporting fleet management (Siddharth et al. 2013), by the prediction of faults (Fan et al. 2015), and (iii) by efficient video surveillance systems (Xu et al. 2011). Concerning railway infrastructure, reliability, availability, maintainability and Safety –RAMS- algorithm could provide security efforts (Park 2013) by means of implementation of defined procedures (Simões 2008). In this last case, also the adoption of technologies such as Ground Penetrating Radar –GPR- monitoring ballast (Zhang 2015), and laser scanner controlling rail surface defects (Xiong 2017) can optimize inspections thus improving monitoring procedures. In this direction data mining could play an important function in the optimization of maintenance procedures (Bastos et al. 2014). In particular Artificial Neural Network –ANN- can be useful to model repairable systems (Rajpal et al. 2006) thus contributing to the intelligent scheduling of maintenances operations. Concerning quarry monitoring systems, Unmanned Aerial Vehicle -UAVs- technologies, combined with reconstruction of topography techniques (Rossi et al. 2017), represent an important way to control the quarry work evolution and consecutively the unsafe areas. By means of the UAVs it is possible the engineering of the risk assessment and of management procedures (Salvini et al. 2017). In all production line processes, image vision techniques can support the risk analysis by automating and integrating a wide range of processes and representations used for defect vision perception (Tushar et al. 2013). The image vision is suitable also for crack detection (Qiao et al. 2013) in quarry conveyor belt. Technology upgrading, RAMS and data mining algorithms can be adopted together for inspections and monitoring applications including quarry machine monitoring.

Key Terms in this Chapter

Infrared Termography: The use of radiometric thermograms to study heat distribution in structures or regions.

Reliability, Availability, Maintainability, and Safety (RAMS): Reliability is a system's ability to perform a specific function and may be given as design reliability or operational reliability; availability is the capability of a system to be considered in a functioning state; maintainability is determined by the simplicity of a system to be repaired or maintained; safety is the requirement not to harm people, the environment, or any other assets during a system's life cycle.

Biometric Recognition: Is the technical approach for body measurements and calculations (it refers to metrics related to human characteristics).

Image Segmentation: Is the process of partitioning a digital image into multiple segments (sets of pixels, also known as super-pixels).

Risk Management: Identification, evaluation, and prioritization of followed by application of resources able to minimize, monitor, and control the probability or impact of unfortunate events and health accidents.

Artificial Neural Network (ANN): The artificial neural network is a data mining framework able to work together and process complex data inputs. Such systems “learn” to perform tasks by considering examples, generally without being programmed with any task-specific rules.

Data Mining: Data mining is the process of discovering patterns in defined data sets involving methods at the intersection of machine learning, statistics, analytics and database systems.

Complete Chapter List

Search this Book: