Fetal Facial Expression Recognition System by Lip Distance Method

The Fetal facial expression is a continuous process related to brain development. 4D ultrasonography scan of fetal showed that expressions are developed at 24 weeks and become complex such as “pain” expressions. Fetal landmarks are not available, so we have to create the landmarks as per requirement because fatal images are not so accurate in detecting face regions like standard face images. Using matlab we create the landmark points of all images available with us. And then it is possible to compare the fetal images with landmarks. In the proposed system, the lip portion is extracted from the foetal face image, and recognition of the unborn baby’s facial expressions is done by designing and implementing a lip distance measurement algorithm for expression classification. This method improves accuracy by up to 20% compared to when the total face is considered for recognition. The landmark points of foetal images are calculated considering the eyes, nose, and lips corner points. A result is compared with the Harris key-point descriptor algorithm and the Gabor feature extraction method. The lip distance method for expression classification is very simple and gives very good results as compared to other methods. This system can be used to find abnormality symptoms in a baby before birth so that the doctor and parents are aware of the baby’s mental state before birth.