Food volume estimation method based on 3D model nesting

Abstract

The present invention discloses a food volume estimation method based on 3D model nesting. The method can restore a simple geometric model corresponding to foods so as to solve the estimated volume ofthe foods. The method comprises the following steps of: selecting a special model to establish a three-dimensional model library, and inputting user hand information; putting the hand and the foods to the same plane for shooting to obtain a three-view image; performing L*a*b color space preprocessing of the image, performing k-means clustering segmentation of a processing result, and obtaining animage segmentation result; performing outer contour matching of the segmentation result and the three views of the model in the three-dimensional model library, and according to the special function,calculating the matching degree of the foods and each model projection map to find out a projection map having the maximum matching degree, wherein the three-dimensional model corresponding to the projection map having the maximum matching degree is a three-dimensional model corresponding to the foods; and taking the hand as a measuring scale to determine the parameters of the matching model so as to estimate the volumes of the foods through a volume calculation formula. The food volume estimation method based on 3D model nesting can simplify the volume reconfiguration complexity to estimatethe effect of food volumes.

Description

technical field [0001] The invention belongs to the technical fields of computer image processing and computer vision, and in particular relates to a food volume estimation algorithm based on nested matching of 3D model outlines suitable for mobile APP applications. Background technique [0002] As early as 1997, the World Health Organization (WTO) had listed obesity as a disease. Studies have shown that at the individual level, people can limit their energy intake from total fat and sugar, and increase their consumption of fruits, vegetables, legumes, whole grains, and nuts. It can be seen that diet control has a great effect on the health management of obese patients. With the popularization of smart phones, it is very convenient to use mobile phones to obtain and process photos of edible food. Users can estimate the volume of food and the composition of various nutrients through the APP, so that users can understand the content of various components of the food they eat,...

AI Technical Summary

Problems solved by technology

However, the currently used method based on contour extraction and model matching nesting requires a lot of user interaction, and the process is cumbersome and not easy to use

Moreover, the calibration process often requires some additional reference objects, which is inconvenient for users to carry and is not conducive to the improvement of user experience.

Images