ENHANCEMENT OF TRIANGLE COORDINATES FOR TRIANGLE FEATURES FOR BETTER CLASSIFICATION

Recently, the triangle features have been applied in digit recognition by adopting the angle as a part of the features. Most of the studies in digit recognition area which applied these features have given impressive results. However, the issue of big gap values that occurred between angles, ratios and gradients has shown a strong impact on the accuracy of the results. Therefore, we introduce our proposed method which is data normalization that has adopted the nature of triangle geometry in order to resolve this issue. Besides, we have applied other techniques, such as Z-score, Minimax and LibSVM function in the experiment. There are four digit datasets used which are HODA, MNIST, IFHCDB and BANGLA. The results of classification have shown that our proposed method has given better results compared to other techniques.