Analisis Dimensi Gambar Terhadap Klasifikasi Batik Indonesia dengan CNN

Abstract

Batik motif classification has gained significant attention due to its cultural significance and practical applications in various fields. This study explores the impact of image dimensions on the classification of batik motifs using Convolutional Neural Networks (CNN). The research investigates how variations in image dimensions affect the accuracy and robustness of CNN-based classification models. Through experimentation with different image resolutions and aspect ratios, the study aims to identify optimal settings for achieving high classification performance. Additionally, it examines the computational efficiency of CNN models under varying image dimensions. The findings contribute to enhancing the understanding of image preprocessing techniques and model optimization strategies for batik motif classification tasks.

Keywords: Batik motif classification, Convolutional Neural Networks (CNN), image dimensions,
classification accuracy, computational efficiency.