<![CDATA[Identification of blood groups by looking at blood clots that have been given reagents is mostly done manually based on direct visual observations on blood that has been dripped with reagents and then determines blood groups based on the clotting of each reagent, where the level of saturation of the eye can affect the results of observations made and the community the general public also does not respect the resemblance. To overcome this problem, an application is made to find blood types using the K-nearest Neighbors method which can work automatically. KNN is a method that classifies new objects based on the training data that is closest to the object, where the distance calculation uses Euclidean Distance. This system is used to identify blood groups A, B, AB and 0. In this study, the system was tested using 80 training data for blood group images which were divided into 20 images for each group A, B, AB and blood group 0. test as many as 40 images consisting of 10 test data for each blood group A, B, AB and blood group 0. The test results for the accuracy of the test on each blood group is A 70%, blood group B 80%, blood type AB 100% and blood group 0 100%. So it can increase the success of the system by 85%.]]>
Copyrights © 2020
