Summary
Activated sludge, used in water treatment, is studied by looking at tiny images under a microscope. The paper shows that new types of artificial intelligence models, called transformer-based models, do a better job than older ones at recognizing patterns in these images. The authors found that making the images too small hurts the model’s accuracy, but making them bigger beyond a certain point does not help much. They also discovered that keeping the whole scene visible in the image is more important than keeping every detail sharp. Finally, the study showed that deep learning methods are more accurate than traditional image analysis for this task.
activated sludgemicroscopic image analysisdeep learningtransformer modelsconvolutional neural networksself-supervised learningimage downsamplingquantitative image analysisclassification accuracyfield of view
Abstract
Microscopic image analysis has long been recognized as a promising approach for monitoring activated sludge. In recent years, deep learning-based image analysis has been increasingly adopted in this field because of its high performance. However, previous studies on microscopic image analysis of activated sludge have rarely explored transformer-based models or self-supervised foundation models and have instead relied on CNNs and supervised ImageNet pretraining. In addition, previous studies often downsampled image sizes, but the effects of downsampling have not been sufficiently investigated, and the relationship between downsampling strategies and image analysis performance remains unclear. Furthermore, no study has quantitatively compared deep learning performance with quantitative image analysis (QIA), which was widely used before the emergence of deep learning. In this study, to examine how model architecture and learning strategies affect performance in microscopic image analysis of activated sludge and to quantitatively determine whether deep learning outperforms QIA, we prepared three types of activated sludge samples, classified their microscopic images, and evaluated classification accuracy. Our results showed that transformer-based architectures and alternative pretraining methods were effective in terms of classification accuracy. Our downsampling analysis showed that using overly small images reduced accuracy, but increasing image size beyond a certain point did not improve it further. In addition, the analysis indicated that, to achieve high classification accuracy, maintaining the field of view was a more effective downsampling strategy than maintaining resolution. Finally, our comparison between deep learning and QIA showed that deep learning outperformed QIA in terms of accuracy.