Automatic classification of the three standard apical views (A4C, A2C and ALX) is essential for the automation of cardiac functional assessment. A multi-stage algorithm for the classification of the apical echocardiogram views is presented. The algorithm employs spatio-temporal feature extraction and supervised dictionary learning approaches to uniquely enhance the classification accuracies of the echocardiograms. A total classification accuracy of 95% is reported with a feasibility to real-time implementation.