Abstarct: Creating geological maps from satellite images can be useful for cost reduction, time saving and providing information of areas that are not easily accessible. For this purpose, detecting lithological boundaries is essential. In most detecting techniques; determination of lithological boundaries is done by classification of an image in to lithological units by spectral variation and then manually draw a boundary. Therefore, identification of lithological boundaries cannot be consistent and is often influenced by the interpreter. The objective of this study is to automatically detect these boundaries using image processing, wavelet transform and classification of the image with support vector machines and clustering. The comparison between the obtained results and the geological map of the study area indicates that implementing these algorithms, the output is adequately accurate. In addition, the lithological units were determined rapidly and the run-time of the algorithm is remarkably low. Another consideration is that mining and the neighboring processing industries cause the low-grade minerals and waste to scatter in the environment which further expands the contamination in the surrounding area. The extracted coal requires washing due to the impurities where waste is produced as low-grade minerals. The composition of the accumulated waste goes through change as they are affected by air and water, causing toxic substance to propagate into the environment. In the following research the detection of contamination zones has also been performed. The results of the study and its comparison with the small scale maps of the study area represent that the algorithm baxsed on clustering, classification and wavelet, benefits from accuracy and speed in creating baxse geological maps and to some extent in mapping the contamination zones.