Classification of motor imagery (MI)-based electroencephalogram (EEG) signals is a key issue for the development of brain-computer interface (BCI) systems. The objective of this study is to develop an algorithm that can distinguish two categories of MI EEG signals. In this paper, we propose a new classification algorithm for two-class MI signals recognition in BCIs. The proposed scheme develops a novel cross-correlation-based feature extractor, which is aided with a logistic regression model. The present method is tested on dataset IVa of BCI Competition III, which contain two-class MI data for five subjects. The performance is objectively computed using a k-fold cross validation (k=10) method on the testing set for each subject. The results of this study are compared with the recently reported eight methods in the literature. The results demonstrate that our proposed method outperforms the eight methods in terms of the average classification accuracy.