Method and device for executing Faster R-CNN neural network algorithm

Abstract

The invention relates to a method and device for executing a Faster R-CNN neural network algorithm. The method comprises the steps that multiple images of different angles of the same food are obtained; a recommendation area of a sample detection is determined by using RPN; the type and frame of a food object in the recommendation area is predicted by using Faster R-CNN; the volume ratio occupiedby each food object is predicted by using Volume R-CNN according to the predicted frame of the food; the volume ratio of different kind of foods is calculated according to the food object type and thefood object volume ratio; the volume ratio of calculated each kind of food multiplies the density of each kind of food to obtain the mass ratio of each kind of food; the total mass ratio of each kindof food multiplies the total mass of the food to obtain the mass of each kind of food; the mass of each kind of food multiplies corresponding nutrient content to obtain food nutrient element content.By means of the method and the device, complicated food of multiple kinds can be measured, and the recognition of food can be more accurate and rapid by adopting an artificial neural network technology and a chip.

Description

technical field [0001] The present invention relates to the technical field of image processing, in particular to a method and device for executing Faster R-CNN neural network operations. Background technique [0002] With the acceleration of the pace of life in modern society and the improvement of people's living standards, people have higher and higher requirements for diet. People no longer care about whether they are full, but whether they eat healthy. However, many people lack sufficient dietary health knowledge, so a device that can intelligently measure food energy and nutritional components is needed to help people eat more reasonably. [0003] One of the existing techniques is to calculate food energy by weight. The metering device mainly consists of the following parts: tray, weight measuring device, and display screen. The weight measuring device is used to measure the weight of the food, and transmits the food weight information to the microcomputer processor...

AI Technical Summary

Problems solved by technology

[0006] The problem with the above technology is that the operation is complicated and the requirements for input photos are high; the side view is prone to food occlusion problems

Using the focal length to predict food parameters such as length, width, and height may have certain deviations for different mobile phones. Using the formula method to calculate food volume is not suitable for irregularly shaped foods, and the calculation results have large errors.

Images