Attention to picture of face, particularly the human face is related to complex information processing in the brain. Humans pay more attention to human faces than other images. The purpose of this study is to verify the existence of particular attention to facial images and categorize the difference between attending to facial and non-facial images through a pair of different pictures as the targets. According to effects of visual stimuli such as color and luminance, the pictures modulated in greyscale (luminance-defined stimuli). Using a psychophysical task, EEG signals according to 10–20 standards in eight channels were recorded from 48 healthy volunteers. After the initial processing, ERP signal were elicited into two classes according to attention to the face and non-face images. In this study, the time window of the N170 component, was considered to extract new time features plus the N170 component; a negative peak in 170 milliseconds after stimulus onset. Optimum features were selected by t test criteria and classification was done by LDA, KNN and SVM classifiers. Validating the results was done by LOO cross validation criteria. Best result was obtained by SVM with 74.44% and was associated with frontal and parietal lobes.