Jump to Main Content
Segmentation of digital rock images using deep convolutional autoencoder networks
- Karimpouli, Sadegh, Tahmasebi, Pejman
- Computers & geosciences 2019 v.126 pp. 142-150
- algorithms, artificial intelligence, color, computers, data collection, neural networks, sandstone
- Segmentation is a critical step in Digital Rock Physics (DRP) as the original images are available in a gray-scale format. Conventional methods often use thresholding to delineate distinct phases and, consequently, watershed algorithm to identify the existing phases. Such methods are based on color contrast, which makes it difficult to automatically differentiate phases with similar colors and intensities. Recently, deep learning and machine learning algorithms have proposed several algorithms working with images, including Convolutional Neural Networks (CNN). Among them, convolutional autoencoder networks have produced accurate results in different applications when various images are available for the training. In this paper, thus, convolutional autoencoder algorithm is implemented to enhance segmentation of digital rock images. However, the bottleneck for applying the CNN algorithms in DRP is the limited available rock images. As an effective data augmentation method, a cross-correlation based simulation was used to increase the necessary dataset in this study. Therefore, using the originally available dataset, namely 20 images from Berea sandstone, a training seed comprising of the manually and semi-manually segmented images was used. Then, the dataset is divided into training, validation and testing groups with a fraction of 80, 10 and 10%, respectively. Next, the produced dataset is given to our stochastic image generator algorithm and 20000 realizations, along with their segmented images, are produced simultaneously. The implemented CNN algorithm was tested for two versions of basic and extended architectures. The results show that the extended network produces results with 96% of categorical accuracy using the designated images in the testing group. Finally, a qualitative comparison with the conventional multiphase segmentation (multi-thresholding) revealed that our results are more accurate and reliable even if very few rock images are available.