Feature selection based Dry-Beans Multiclass classification with Optimized Deep Neural Network

Varieties of dry beans are available in the world. Therefore, the quality of seed is necessity to maintain at the time of harvesting and production as well. Hence, the seed classification based on it dimension and size is necessary. The objective of this research is to provide an automatic method to classify dry bean based on it dimension and size. The classification technique helps farmers and industrialists to classify good and bad quality seeds. A dataset of 13611 grains images is used here to distinguish seven types of dry bean. High-resolution images are used to extract features such as dimension and size. The feature consists of 16 dimensions and 4 shape that further normalized using principal component analysis. The proposed technique is compared with the various state of the art techniques and outperforms. The proposed techniques attain 99.5% classification accuracy.