Non-invasive and non-contact automatic jaundice detection of infants based on random forest

Jaundice is a common phenomenon in neonates and significant cause of morbidity mortality. Early detection jaundice still challenge. Moreover, existing invasive techniques are stressful, painful for the newborn, non-invasive devices expensive. Therefore, we investigate characteristics non-contact neonatal system based on skin colour analysis machine learning using graphical user interface (GUI). First, automatically selected region interest (ROI) from image an infant captured by digital camera. Then, ROI was analysed both RGB YCbCr spaces. Finally, algorithm random forest (RF) incorporated to classify or normal determine whether neonate requires treatment not. The experimental result demonstrates that proposed has potential as technique clinical application.