The research area of the indoor scene recognition has attracted the various scientists and engineers across the globe, which includes the neuroscientists, electronics engineers, robotic engineers, digital image experts, camera developers and manufacturers for the purpose of application designing in the fields of the computer vision, vision based communications and the access control systems. The indoor scene recognition methods require the inclusion of the various methods in the computer vision, image processing and feature recognition for the scene recognition by identifying the category of the input image by comparing it against the given training databases by the means of the feature descriptor (popularly based upon the color or low level features) and the classification algorithm. The indoor scene classification algorithms require the number of the computations and feature transformations along with the normalization and automatic categorization. In this thesis, the multi-category dataset has been incorporated with the robust feature descriptor using the scale invariant feature transform (SIFT) along with the multi-category enabled support vector machine (mSVM).