The aim of this article is to compare the performance of three image processing techniques for recognizing the tanker in land chemical emergencies. The best algorithm will be part of a future electronic platform for attending chemical emergencies. Such platform should be portable and should allow its installation in different devices, i.e. drones. To be realistic, the images were obtained with a drone. These images were processed via three methods: Hotelling transform, Hu moment invariants, and Signature transform (Distance vs. Angle). The results showed that Hu was the most accurate method, obtaining 86% of accuracy in the tests; followed by Signature with 71%; and the less reliable in this case was Hotelling, with only a 23%. The future platform will not be autonomous, so it will always be supervised by an operator and only help them in the decision-making process. Since this is a small piece of a bigger research, it still has a lot of future work. Data base will be increased, and future improvements in the performance of the algorithm will be tested. Nevertheless, the objective of this paper was achieved as the recognition algorithm worked, with enough accuracy, for the purpose it was aimed to.