PSDNet: Determination of Particle Size Distributions Using Synthetic Soil Images and Convolutional Neural Networks
This work addresses a domain-specific problem in soil analysis by providing an automated image-based method for particle size distribution, but it is incremental as it applies existing CNN models to synthetic data.
The paper tackled the problem of determining grain size distributions from images of granular materials using convolutional neural networks, achieving root-mean-square errors as low as 1.8% for predicting mass percentages on sieves with a coefficient of determination of 0.99.
This project aimed to determine the grain size distribution of granular materials from images using convolutional neural networks. The application of ConvNet and pretrained ConvNet models, including AlexNet, SqueezeNet, GoogLeNet, InceptionV3, DenseNet201, MobileNetV2, ResNet18, ResNet50, ResNet101, Xception, InceptionResNetV2, ShuffleNet, and NASNetMobile was studied. Synthetic images of granular materials created with the discrete element code YADE were used. All the models were trained and verified with grayscale and color band datasets with image sizes ranging from 32 to 160 pixels. The proposed ConvNet model predicts the percentages of mass retained on the finest sieve, coarsest sieve, and all sieves with root-mean-square errors of 1.8 %, 3.3 %, and 2.8 %, respectively, and a coefficient of determination of 0.99. For pretrained networks, root-mean-square errors of 2.4 % and 2.8 % were obtained for the finest sieve with feature extraction and transfer learning models, respectively.