A Brain-Machine Interface (BMI) is an intelligent system that records electroencephalogram (EEG) signals from human. The system records the brain perception signal while the subject is made to visualize a picture and later pre-processed. After pre-processing, the band energy features are extracted using a simple feature extraction technique. The extracted features are then associated with the respective visualization task perception and a neural network model is developed. The system has the capability of recording and processing the EEG perception signals and identifying the correct perception signal. Once the network identifies the perception signal, it is then converted into an equivalent voice signal and played through an audio speaker.