ROTATION-GAMMA CORRECTION AUGMENTATION ON CNN-DENSE BLOCK FOR SOIL IMAGE CLASSIFICATION

• DOI: 10.35784/acs-2023-27
• Publication Date: 2023-10-26T07:38:53Z
• Authors: Sri INDRA MAIYANTI; Anita DESIANI; Syafrina LAMIN; P PUSPITAHATI; Muhammad ARHAMI; Nuni GOFAR; Destika CAHYANA
• Keywords: Contextual image classification

Soil is a solid-particle that covers the earth's surface. Soils can be classified based their color. The color an indication of soil properties and conditions. image classification requires high accuracy caution. CNN works well on classification, but large amount data. Augmentation one technique to overcome data needs like rotation improving contrast. Rotation movement rotating position randomly various degrees. Gamma Correction method improve by decreasing or increasing augmentation increase training from 156 2500 images not referred taxonomy system such as Entisols Histosols it used arbitrary simple Unfortunately, weakness vanishing exploded gradients. Another Deep learning gradients dense blocks. This study proposes combination CNN-Dense block where in Gamma-correction techniques block. able give excellent results, all performances accuracy, precisions, recall F1-Score are above 90%. robust use