Infant Mortality Rate (IMR) is the predominant problem in medical, especially in developing countries like Indonesia. One cause of IMR increasing is diagnosis mistake of infant pain. Baby's communication have not perfect, it is leading to high risk of diagnosis mistake when the baby feels pain, such as post-surgery conditions. This problem encourages research on the pain identification of infant facial expressions. The purpose of this study is to analyze change or difference in infant facial expressions. This research is expected to provide the appropriate parameters in order to contribute the accuracy of infant diagnosis. The alteration of infant's facial expressions is analyzed using several methods as appropriate, facial feature detection and extraction take advantages of Ganglion Cells working principles, which nCRF (non-Classical Receptive Field) mechanism and inhibition and disinhibition are complementary in weak feature detection. There are several ways of infant pain identification in medical, Neonate Facial Coding System (NFCS) is one which focuses in face. With several points of NFCS scale is expected to get more accuracy in the pain identification. This study is a combination of image processing and medical methods that can be more flexible in future development.