Enhancement of thermal camera images

All objects emits infrared radiation in any amount according to their structure. Since thermal cameras work with different optical spectra, they provide a heat map by detecting this infrared heat, emitted by objects, which human cannot see. Thus, it is possible to decompose the objects by benefiting from the temperature difference and the living/lifeless objects can be identified separately. However, it is detected many times that the thermal cameras are effected by the bad weather conditions and noise, apart from this it is recognized that their brightness level is low/high and their contrast settings are not proper. Especially like a black colored umbrella, substances blocking the heat emitted from objects make the image detection more difficult. Because of all these negative reasons, it is difficult to distinguish the objects accurately and select them clearly. In this thesis, thermal images have been enhanced by using derivative operations and Fourier transform. Through this technique, the thermal objects on the image have been made clear. Furthermore, the distortion, lack of contrast on the image and the other objects have been clarified. The experimental result show that the developed method to recover thermal image is extremely succesful. Through this method, various problems coming out of illusions faced in the practice will be prevented. Key Words: Image processing, Thermal imaging, Image enhancement, Thermography