This work addresses the problem of semantic foggy scene understanding (SFSU). Although extensive research has been performed on image dehazing and on semantic scene understanding with clear-weather images, little attention has been paid to SFSU. Due to the difficulty of collecting and annotating foggy images, we choose to generate synthetic fog on real images that depict clear-weather outdoor scenes, and then leverage these partially synthetic data for SFSU by employing state-of-the-art convolutional neural networks (CNN). In particular, a complete pipeline to add synthetic fog to real, clear-weather images using incomplete depth information is developed. We apply our fog synthesis on the Cityscapes dataset and generate Foggy Cityscapes with 20550 images. SFSU is tackled in two ways: 1) with typical supervised learning, and 2) with a novel type of semi-supervised learning, which combines 1) with an unsupervised supervision transfer from clear-weather images to their synthetic foggy counterparts. In addition, we carefully study the usefulness of image dehazing for SFSU. For evaluation, we present Foggy Driving, a dataset with 101 real-world images depicting foggy driving scenes, which come with ground truth annotations for semantic segmentation and object detection. Extensive experiments show that 1) supervised learning with our synthetic data significantly improves the performance of state-of-the-art CNN for SFSU on Foggy Driving; 2) our semi-supervised learning strategy further improves performance; and 3) image dehazing marginally advances SFSU with our learning strategy. The datasets, models and code are made publicly available.

Semantic Foggy Scene Understanding with Synthetic Data

This work addresses the problem of semantic foggy scene understanding (SFSU). Although extensive research has been performed on image dehazing and on semantic scene understanding with clear-weather images, little attention has been paid to SFSU. Due to the difficulty of collecting and annotating foggy images, we choose to generate synthetic fog on real images that depict clear-weather outdoor scenes, and then leverage these partially synthetic data for SFSU by employing state-of-the-art convolutional neural networks (CNN). In particular, a complete pipeline to add synthetic fog to real, clear-weather images using incomplete depth information is developed. We apply our fog synthesis on the Cityscapes dataset and generate Foggy Cityscapes with 20,550 images. SFSU is tackled in two ways: (1) with typical supervised learning, and (2) with a novel type of semi-supervised learning, which combines (1) with an unsupervised supervision transfer from clear-weather images to their synthetic foggy counterparts. In addition, we carefully study the usefulness of image dehazing for SFSU. For evaluation, we present Foggy Driving, a dataset with 101 real-world images depicting foggy driving scenes, which come with ground truth annotations for semantic segmentation and object detection. Extensive experiments show that (1) supervised learning with our synthetic data significantly improves the performance of state-of-the-art CNN for SFSU on Foggy Driving; (2) our semi-supervised learning strategy further improves performance; and (3) image dehazing marginally advances SFSU with our learning strategy. The datasets, models and code are made publicly available.

Pipelines are widely used for the transportation of hydrocarbon fluids over millions of miles all over the world. The structures of the pipelines are designed to withstand several environmental loading conditions to ensure safe and reliable distribution from point of production to the shore or distribution depot. However, leaks in pipeline networks are one of the major causes of innumerable losses in pipeline operators and nature. Incidents of pipeline failure can result in serious ecological disasters, human casualties and financial loss. In order to avoid such menace and maintain safe and reliable pipeline infrastructure, substantial research efforts have been devoted to implementing pipeline leak detection and localisation using different approaches. This paper discusses pipeline leakage detection technologies and summarises the state-of-the-art achievements. Different leakage detection and localisation in pipeline systems are reviewed and their strengths and weaknesses are highlighted. Comparative performance analysis is performed to provide a guide in determining which leak detection method is appropriate for particular operating settings. In addition, research gaps and open issues for development of reliable pipeline leakage detection systems are discussed.

Recent Advances in Pipeline Monitoring and Oil Leakage Detection Technologies: Principles and Approaches

Pipeline corrosion and leakage monitoring based on the distributed optical fiber sensing technology

With the development of natural gas transportation systems, major accidents can result from internal gas leaks in pipelines that transport high-pressure gases. Leaks in pipelines that carry natural gas result in enormous financial loss to the industry and affect public health. Hence, leak detection and localization is a major concern for researchers studying pipeline systems. To ensure the safety and improve the efficiency of pipeline emergency repair, a high-pressure and long-distance circular pipe leakage simulation platform is designed and established by similarity analysis with a field transmission pipeline, and an integrated leakage detection and localization model for gas pipelines is proposed. Given that the spread velocity of acoustic waves in pipelines is related to the properties of the medium, such as pressure, density, specific heat, and so on, this paper proposes a modified acoustic velocity and location formula. An improved wavelet double-threshold de-noising optimization method is also proposed to address the original acoustic wave signal collected by the test platform. Finally, the least squares support vector machine (LS-SVM) method is applied to determine the leakage degree and operation condition. Experimental results show that the integrated model can enhance the accuracy and precision of pipeline leakage detection and localization.

Integrated leakage detection and localization model for gas pipelines based on the acoustic wave method

Gas leaks can cause major incidents resulting in both human injuries and financial losses. To avoid such situations, a considerable amount of effort has been devoted to the development of reliable techniques for detecting gas leakage. As knowing about the existence of a leak is not always enough to launch a corrective action, some of the leak detection techniques were designed to allow the possibility of locating the leak. The main purpose of this paper is to identify the state-of-the-art in leak detection and localization methods. Additionally we evaluate the capabilities of these techniques in order to identify the advantages and disadvantages of using each leak detection solution.

https://www.aut.upt.ro/~pal-stefan.murvay/papers/survey_gas_leak_detection_localization_techniques.pdf

A survey on gas leak detection and localization techniques

The latest researches in Automotive field concentrate on increasing the security and safety of the passengers. The automotive companies invest a lot in testing their products before delivery, that is why the cars are very safety from this point of view. Most of the accidents nowadays happen because of the driver, caused by bad visibility, hurry or inattention. In this paper, a system of detecting visibility in foggy environment was developed which gives the driver a feedback on the visibility distance and an advice to adapt the speed at a value at which the car can be stopped in safety conditions if an obstacle appears on the road. The system can be used on the highways and after some measurements a feedback is displayed on the display panels from the highway or sent on driver's smart-phones.

Visibility Detection in Foggy Environment

The manufacturing industry is continuously researching and developing strategies and solutions to increase product quality and to decrease production time and costs The approach is always targeting more automated, traceable, and supervised production, minimizing the impact of the human factor In the automotive industry, the Electronic Control Unit (ECU) manufacturing ends with complex testing, the End-of-Line (EoL) products being afterwards sent to client companies This paper proposes an image-processing-based low-cost fault detection (IP-LC-FD) solution for the EoL ECUs, aiming for high-quality and fast detection The IP-LC-FD solution approaches the problem of determining, on the manufacturing line, the correct mounting of the pins in the locations of each connector of the ECU module, respectively, other defects as missing or extra pins, damaged clips, or surface cracks The IP-LC-FD system is a hardware-software structure, based on Raspberry Pi microcomputers, Pi cameras, respectively, Python and OpenCV environments This paper presents the two main stages of the research, the experimental model, and the prototype The rapid integration into the production line represented an important goal, meaning the accomplishment of the specific hard acceptance requirements regarding both performance and functionality The solution was implemented and tested as an experimental model and prototype in a real industrial environment, proving excellent results

Image-Processing-Based Low-Cost Fault Detection Solution for End-of-Line ECUs in Automotive Manufacturing.

The Internet of Things means a connected world. When reaching the industrial level, the automation equipment present a large life cycle and the systems are many times conceived using proprietary protocols. Bringing interoperability to local automation is often a difficult task and changes in the functional structures, particularly in the case of critical infrastructures like the water industry, have to be of the lowest possible magnitude. The paper proposes a low-cost middleware OPC UA wrapping structure in order to assure interoperability of the local automation. The solution is also offering the possibility to monitor and control the local system, to create a local historian and to further implement control algorithms for existing structures or to optimize existing tags packaging for further Supervisory Control and Data Acquisition (SCADA) integration without modifying the local software. The developments were tested in real scenarios, a case study in the water industry being presented.

Achieving interoperability using low-cost middleware OPC UA wrapping structure. Case study in the water industry

The increased number of cyber threats against the Supervisory Control and Data Acquisition (SCADA) and automation systems in the Industrial-Internet-of-Things (IIoT) and Industry 4.0 era has raised concerns in respect to the importance of securing critical infrastructures and manufacturing plants. The evolution towards interconnection and interoperability has expanded the vulnerabilities of these systems, especially in the context of the widely spread legacy standard protocols, by exposing the data to the outside network. After gaining access to the system data by launching a variety of attacks, an intruder can cause severe damage to the industrial process in place. Hence, this paper attempts to respond to the security issue caused by legacy structures using insecure communication protocols (e.g., Modbus TCP, DNP3, S7), presenting a different perspective focused on the capabilities of a trusted platform module (TPM). Furthermore, the intent is to assure the authenticity of the data transmitted between two entities on the same (horizontal interoperation) or different (vertical interoperation) hierarchical levels communicating through Modbus TCP protocol based on functionalities obtained by integrating trusted platform modules. From the experimental results perspective, the paper aims to show the advantages of integrating TPMs in automation/SCADA systems in terms of security. Two methods are proposed in order to assure the authenticity of the messages which are transmitted, respectively the study presents the measurements related to the increased time latency introduced due to the proposed concept.

/pdf/cryptographic-considerations-for-automation-and-scada-1llg8hg3vi.pdf

Ioan Silea

Papers

A survey on gas leak detection and localization techniques

Visibility Detection in Foggy Environment

Image-Processing-Based Low-Cost Fault Detection Solution for End-of-Line ECUs in Automotive Manufacturing.

Achieving interoperability using low-cost middleware OPC UA wrapping structure. Case study in the water industry

Cryptographic Considerations for Automation and SCADA Systems Using Trusted Platform Modules.