Abstract
Data fusion on remote sensing refers to the process of combining data from multiple Sensors to produce a dataset. The resulted data set contains more detailed information then each of the individual sources also data from multiple sensors can significantly improve the accuracy and interpretation of features. Pixel-level, Feature-level and Decision-level are three common levels of data fusion. Here, we summarized the proposal work, to increasing the testing samples and reduce the computational complexity of processing input image. Generally, spectral resolution will give more about information of the spectrum of the image and LiDAR gives the light intensity if the image. The Novelty of this work to reduce the training samples and achieve better classification accuracy in urban places. This work, concentrates on the subsequent intentions: (i) To enhance the classification precision in HSI and LiDAR images (ii) To lessen the assorted pixels in suburban constructions that are encircled by tiny trees (iii) To reduce the comparable pixels of streets and parking areas. Here, 15 separate classes were categorized which are significant for the expansion in urban areas.