This paper proposes the feasible method for American Sign Language recognition. We design the glove with 6 different color markers, and the algorithm which has three main processes; firstly, marker detection by using Circle Hough Transform, next is calculation all feasible triangle areas from 3D coordinate point that is novel feature, and process of classification that we use feedforward backpropagation of Artificial Neural Network. The experiment performs the accuracy of the algorithm. The experimental result illustrates the average of accuracy is 95 percent that is high performance