Pierre Sarnow,Sean M. McNamara,Thomas M. Moretto, JR.
申请号：
US15482519
公开号：
US20170209090A1
申请日:
2017.04.07
申请国别(地区):
US
年份:
2017
代理人:
摘要：
Provided is a non-invasive system and method of determining human body fat based on image processing. The method includes receiving at least one ultrasound scan image of at least a portion of a skin layer as disposed above one or more additional tissue layers, the skin layer defining a horizontal axis and the image provided by a plurality of pixels. The method continues by horizontally blurring the pixels of the image and thresholding the pixels of the image to provide a binary image having a plurality of elements of different sizes. The method continues with morphing the structural elements of the binary image to remove small elements and connect large elements. With this resulting image, the method distinguishes a body fat layer from the remaining elements and determines the body fat layer thickness. An associated system is also disclosed.
