Person Segmentation in Human Height Measurement Using Bodypix

Abstract

Image processing can be used to measure height. Most research in this measurement uses a place adjusted to the green screen background to facilitate image segmentation. Using a green screen background can facilitate the image segmentation process, but using the space in question is inefficient. In this paper, a height measurement experiment is carried out using the body pix module to segment images of human height. The body pix module is expected to make the image processing process more efficient and inexpensive. From the study results, the best height calculation accuracy using the body pix module is 99.927059%. This experiment does not require any additional devices on the background of the human object, so it is cheaper and more efficient.