Down syndrome is a genetic disorder and is caused by an extra copy of the 21st chromosome. This extra genetic material causes the physical and developmental characteristics associated with Down syndrome. It is known that the quality of life of individuals can be increased with developing robotic technology. This situation can be used to customize for individuals with Down syndrome, to enable them to adapt to social life more easily and to reveal their potential. The aim of the paper is to detect individuals with Down syndrome and to enable approaching these individuals differently. The outputs of this paper can be used for some devices or equipment such as humanoid robots produced for people with down syndrome. Therefore, image processing models have been developed with some methods. We tried to determine which method achieved the highest success rate of the models we developed using pre-trained models. We achieved very high success rates of around 90%.