Start up the automatic measurement of urinary volume by image processing on Raspberry Pi

Abstract

Research on measurement and management of urinary volume in the toilet has been conducted mainly from the demand in medical institutions and welfare facilities. We have proposed liquid flow estimation using image processing as a convenient and hygienic urinay volume measurement method. In this method, the streakline in the image is parabolic approximated and its initial radiation velocity is derived to obtain the volume by the multiple cylindrical model. Although it seems that the introduction have a significant hurdle to overcome when performing such image processing, in fact it is much-anticipated sufficient practical measurement accuracy even with an inexpensive single-board computer like Raspberry Pi and a general-purpose image processing library. In this research, we introduce the implementation by using Raspberry Pi and describe the image processing algorithm which can improve accuracy by specifying the error factor happened in the cheap camera system.