In this paper a computerized scheme for automatic detection of tumors in brain is examined. Diagnosis of these lesions at the early stage is a very difficult task in normal brain images. The algorithm incorporates steps for preprocessing, feature extraction and classification using brain tumor detection. This paper proposes a supervised machine learning algorithm for detection of tumor. A feature extraction methodology is used to extract the Gabor texture features of the abnormal brain tissues and normal brain tissues prior to classification. Then support vector machine classifier is applied at the end to determine whether the given input data is tumor or non tumor. The detection performance is evaluated using Receiver Operating Characteristic curves. The result shows significantly improves the classification accuracy. © Springer International Publishing Switzerland 2015.