Motor neuron disease (MND) is a condition where voluntary muscle movements of the patient stop functioning due to progressively damage of motor neurons resulting paralyzed the patient. One of the solution of MND is motor imagery (MI) based brain-computer interface (BCI) which acts as an assistive device for motor disabled people. But it has limited applications due to its lower classification performance. To improve it, this paper introduces second order difference plot (SODP) for the detection of various MI activities. First, filter bank technique was implemented to the signals and set of multiple sub-bands were generated. In order to study MI activities effectively, SODP was applied to each sub-band and area of ellipse was calculated. The feature (area of ellipse) of all sub-bands were combined and the significant features (p <0.05) were extracted using one-way analysis of variance (ANOVA). These significant features were fed into multi-class support vector machine (SVM) for decoding MI activities. The Proposed method and classifier were tested on BCI competition 2008 MI dataset-II-a. The performance of the proposed method was evaluated in term of Cohen's kappa coefficient (K). Results show that the SVM improved the mean value of kappa (K=0.62) and outperformed the existing methods reported in the literature. © 2019 IEEE.