Hyperspectral imagery is a faraway sensing fact that gives one-of-a-kind spectral reflectance measurements across the electromagnetic spectrum. This form of imagery has various uses for monitoring the environment and tracking flora, wooded area cowl, and urban increase. However, successful analysis of hyperspectral imagery requires correct item-orientated classification, which may be challenging and time-consuming to acquire with traditional methods. This paper affords a choice tree-based method to the object-oriented type of hyperspectral imagery. This technique uses statistical class strategies and, blended with professional information, picks out the numerous objects described in the imagery. The proposed approach has been tested on hyperspectral imagery from exceptional environments, and the consequences reveal high accuracy and occasional processing times. This approach can offer a more accurate object-oriented class of hyperspectral imagery than conventional strategies.