The indoor scene recognition is the process of recognizing the scene in focus, which can be either an office or home. The indoor scene recognition has many real-time applications, which includes the automatic camera mode selection, robotic mobility, automatic CCTV coverage and alert setting based upon region visiting frequency, etc. The proposed model is designed for the indoor scene recognition based upon the multiple features, which primarily involves the combination of textural and color-pattern based features. The scene recognition is entirely based upon the pattern & texture recognition, which includes the speeded up robust features (SURF), HOG and scale invariant feature transformation (SIFT) features to differentiate the different indoor scenes. The proposed model has coupled with Classify Neural Network (CNN) classification algorithm. The proposed model has achieved the perfect accuracy of nearly 94%, which has outperformed all other data models with individual features of SIFT, SURF or HOG. This shows the effectiveness of the proposed model based upon multi-feature combination in comparison with the individual feature based applications.