Image Processing Method for Cargo Container Identification in a Stack Within the Cargo Temperature Control and Fire Safety System on Container Ships

Abstract

Abstract The current research is focused on the identification of cargo containers in a stack from their images in the infrared and visible spectra, in order to locate the container-origin of ignition within the cargo temperature control and fire safety system. The relevance of the topic is reinforced by the functional requirements for shipboard safety, which are embodied in Chapter II-2 of the Safety of Life at Sea (SOLAS) Convention, and demanded by the necessity of enhancing safety measures during cargo transportation by the world container fleet. The thermal imager’s field of view (FOV) and the coordinate dependencies between the object and its image have been studied and modelled, and an algorithm for fire detection has been defined within the scope of the current research in connection with the containers within the camera’s FOV. A corresponding verification has been carried out by means of simulation modelling using the Unity and C# programming language capabilities.