Doctoral thesis – A contribution to increasing the efficiency of the Ship fire detection system by implementing new electronic and computer technologies – examines an approach to the increasing ship's safety in a case of fire. The traditional way of fire protection involves three areas: structural fire protection, fire detection and fire fighting. More than half of a ship's fires occur in the area of the ship's engine room, which is why the emphasis of the research is focused precisely in this area. Damage and losses incurred due to fire on ships in recent decades have accelerated the development of new techniques, technologies and innovations to increase the safety and reliability of fire detection systems. Early detection plays a crucial role in fire fighting and prevention of fire spreading. The development of fire alarm technology has been driven by a number of fire incidents which have caused huge material losses, with an even greater human toll. The current development of fire alarm detectors with the integration of technology of fire detection system with micro-electronic and information technology has created a high degree of system autonomy, while increasing reliability and availability. At the same time, the increasing demand for improved technology of early detection of fire and its prevention has developed advanced detection systems whose control algorithms have the task of intelligent processing, distinguishing real from false signals of smoke or flames. Through usage of a service offered by the new technology, the ship operator aims to gain quality crew protection, and the protection of the ship's inventory, as well as reduction of expenses that could appear due to an accident caused by a fire. With an overview of the latest techniques and technologies for early detection of fires on ships, this doctoral thesis shows the improvement of the characteristics of new systems and components of detection systems, analyzing their features, advantages and applications. The results of a research has shown that an implementation of a computer vision based fire detection can improve current classical fire detection system, by reaching greater speed of fire detection, thus faster fire containment and fire suppression.