The fundamental policy of marble industries is to establish sustainable high-quality products in a standardized manner. Identification and classification of different types of marbles is a critical task that is usually carried out by human experts. However, marble quality classification by humans can be time-consuming, error-prone, inconsistent, and subjective. Automated and computerized approaches are required to obtain faster, more reliable, and less subjective results. In this study, a deep learning model is developed to perform multi-classification of marble slab images with six different quality types. Blur filter, 5 ✕ 5 low-pass 2D linear separable convolution filter using Gaussian kernel, and erosion filter were applied to the images for data augmentation, and a special convolutional neural network (CNN) architecture was designed and implemented. It has been observed that the data augmentation approach for marble image samples has significantly improved the accuracy of the CNN model ranging between 0.922 and 0.961.

 Use of convolutional neural networks for multi-classification  Marble quality classification with slab images  Data augmentation with image processing filtersIMarble quality classification by human experts can be time-consuming, error-prone, unreliable, and subjective. Automated and computerized methods are needed to obtain more reliable, faster, and less subjective results. In this study, a deep learning model is developed in order to perform multi-classification of marble slab images with six different quality types. Some special image pre-processing operations were applied to the images for data augmentation and a special convolutional neural network (CNN) architecture was designed and implemented. It has been observed that the data augmentation approach for marble image samples has significantly improved the accuracy of the CNN model. Purpose:This study aims to propose a novel marble quality classification model that could provide accuracy scores at least as good as human experts, which might be implemented and used in the marble industry as autonomous agents that could minimize human intervention and human workforce requirements in the near future. To the best of our knowledge, the use of CNN for marble quality multi-classification and image preprocessing and data augmentation techniques for marble quality classification is known to be a novel approach and model in the related literature. Theory and Methods:The original dataset consisted of 2100 marble images with six different classes. Image blurring was achieved by convolving the image with a low-pass filter 4x4 kernel. After that, 2D linear separable filter was applied. After these data augmentation operations, the marble image dataset size was increased up to 6300. Batch normalization was used for the input data with a batch size of 16 within CNN. Two convolution layers with 64 and 128 filters of (3x3) sizes were used. A fully connected layer was used where 139392 nodes were fully connected with 512 nodes and a dropout rate of 0.50. CNN model was trained for 30 epochs and Adam is used with an initial learning rate of 0.001. Results:The average accuracy scores obtained by our CNN model using the augmented dataset with 6100 images were 0.922 (test) and 0.961 (ten-folds cross-validation), which are both much higher than the results obtained by other algorithms. In addition, these results are considered to be at least as good as (or even exceeding) the human experts' manual classification. The novelty of our study and its contributions can be summarized as follows; the use of a specifically designed convolutional neural network for marble quality classification, the use of specific two image processing methods (Blur filter and 2D linear separable filter) for data augmentation, the delivery of highly successful results without overfitting problems. Conclusion:It can be concluded that this CNN and data augmentation model can be implemented and used in marble companies as a fully automated and computerized alternative for manual operations carried out by hu...