Activated sludge (AS) is a type of process which is commonly used for the treatment of sewage and industrial wastewater. In this treatment process, the settling of the sludge flocs is important to ensure the normal functioning of the system, while sludge bulking has become a common and long-term problem that greatly affects floc settleability. Thus, methods based on image processing and analysis are introduced for monitoring AS wastewater treatment plants. However, the effectiveness of using image processing methods heavily depends on the performance of segmentation algorithms. The AS wastewater plant can be monitored through microscopic images of the flocs and filaments. Water samples are taken from the aeration tank of the wastewater plants and then observed using bright field and phase-contrast microscopy to compare the segmentation accuracy at different magnifications i.e., 4x, 10x, 20x, 40x. In this paper, three methods to segment and quantify the flocs in bright field and phase-contrast microscopy images have been analyzed. The first method is image segmentation using Bradley local thresholding method, the second method is texture segmentation using range filtering and Otsu’s thresholding and the third method is Gaussian Mixture Method based segmentation. The experimental results show that Gaussian Mixture Model Method gives the best segmentation accuracy for bright-field microscopy and 10x magnification gives the best results.