Identifying bank cheques is one of the important task occurring in cash transactions, as banks receives huge number of different bank cheques every day from various account holders. For country like India, it is big hurdle. Identifying bank cheques, manually can be done but It is very time consuming. This paper presents the novel method for identification and classification process which can be efficiently done. The proposed approach works on a specific features of the Indian bank cheque and classify according to different banks. The recognition system is having three parts. First, The scanned image is processed by reducing  size and extracting its features by considering 18 color features that is RGB(Red Green Blue), HSV(Hue Saturation Value). Second, Logo detection using Speeded Up Robust Features (SURF) point matching algorithm and extract region properties of logo and Third Classification of different bank cheques using neural network classifier. The Bank Cheque Identification and Classification system has given  91.43 accuracy.