A Simple Image Analysis Algorithm For Evaluation Of Extendedfilaments Length Based On The Enhanced Digitized Image

摘要

An automated image analysis procedure was developed to determine extended filaments length and floc area to evaluate settling characteristics of activated sludge. Digitized image obtained by Gram staining granted the first step of algorithm, segmentation, to be extremely clear and simple. The image analysis work could become more accurate and less time consuming one to be required only 1 minute of operation time per image. Filamentous bulking phenomenon of the biological sludge was induced in a laboratory-scale bioreactor system under an extreme operational condition, and the non-bulking and filamentous bulking sludge samples captured were examined with image analysis as well as traditional settling test. There existed a linear relationship between DSVI (Diluted Sludge Volume Index) and the extended filaments length. Similar results were also obtained with the extended filaments per floc area. The arithmetic mean extended filaments length showed an excellent linearity with the DSVI. It is believed that the image analysis algorithm developed in this study can be utilized for estimation of the extended filaments length and for evaluation of the settling characteristics of activated sludge effectively.