Eye tracking is a technique used for driver assistance systems since human eyes contain information about the attention level of the driver. Identifying the region of eyes on face will help to detect drowsiness and fatigue of drivers. The existing eye tracking algorithms are sensitive to the lighting conditions. This has a tendency to significantly limit the scope in the automobile applications. This paper focuses on the recognition of eyes even in poor lighting conditions. We have used gamma correction algorithm to enhance the image and have applied Haar classifiers for eye detection. The proposed model can process low quality images with high detection rates. These kind of algorithms or techniques are needed to incorporate in the driver assistance systems in the vehicles to detect the fatigue of drivers. It is observed that the proposed algorithm is giving better results for images on poor lighting conditions. © 2018, Institute of Advanced Scientific Research, Inc.. All rights reserved.