Significant growth of the population in cities in the last few decades requires close monitoring of urban change. Monitoring can be applied as an influential factor in the field of urban management and planning. It also helps to estimate the amount of damage caused by natural disasters such as Earthquakes, floods, and fires. Recent improvement of the quality of satellite images and the development of machine learning methods have made the change monitoring algorithms more accurate and faster than before. In this article, buildings change is monitored using the U-Net++ deep learning model and Onera satellite change detection dataset by means of exploiting input data combinations in different approaches in the arrangement of spectral bands, remote sensing indices and extracted features. The feature selection is to reduce the dimensionality of the input data to the network. Unlike ordinary feature extraction methods that normally extract high-level features, the feature extraction method used in this study is based on the level of complexity of the data. The data combinations are then used as input data to the U-Net++ deep learning model. The results show that the use of spectral indices can improve the performance of the model. By applying the feature extraction process to reduce the input data dimensionality, the training time of the model was reduced and the network convergence accelerated considerably. However, this considerable reduction in processing time did not sensibly affect the final accuracy of the results.