Fruit and vegetable size identification method and device, electronic equipment and computer readable medium

Abstract

Disclosed is a fruit and vegetable size identification method comprises the following steps: inputting a picture comprising a fruit and vegetable image and a reference object image into a neural network; identifying image boundaries of fruits and vegetables, image boundaries of reference objects and types of the reference objects in the pictures by utilizing a neural network; acquiring the actualsize of the reference objects from a reference object actual size database according to the types of the reference objects; and calculating the actual sizes of the fruits and vegetables according to the actual sizes of the reference objects, the image boundaries of the reference objects and the image boundaries of the fruits and vegetables. The neural network is formed by training based on a training sample, collects image information of a reference object and image information of fruits and vegetables to generate the training sample, and obtains actual size information of the reference objectto establish the reference object actual size database. According to the fruit and vegetable size identification method, the fruit diameter can be identified by using a mobile phone to take photos bytaking common objects as reference objects in a natural illumination environment, so that the operation is convenient, and the application range is wider.

Description

technical field [0001] The present invention relates to the field of computer vision recognition, in particular to a method, device, electronic device and computer-readable medium for recognizing the size of fruits and vegetables based on machine vision. Background technique [0002] In the process of selling agricultural products, agricultural products of different sizes are generally sold at different prices. For example, apples can be divided into three sizes: 70 (mm), 80 (mm) and 90 (mm) according to the fruit diameter. Generally, the larger the fruit diameter, the higher the price. Currently, machine vision is increasingly used to automatically identify the size of the fruit diameter to reduce hassle and increase efficiency. [0003] In a known apple grading method, a fixed camera is first used to capture the picture of the apple on the conveyor belt, the edge of the apple is separated by a related algorithm, and the number of pixels occupied by the length and width of...

AI Technical Summary

Problems solved by technology

[0005] In the above apple grading method, the edge separation effect of the algorithm is poor, and the distance between the camera and the conveyor belt or the shooting focal length may change during the measurement process, resulting in a change in the number of pixels occupied by the apple, thereby affecting the measurement of the fruit diameter. Defects

[0006] In the above another fruit grading technology, the conveyor belt needs to be used as a monochromatic background, so the fruit cannot be detected in the background of the natural scene, and the calculation of the fruit diameter depends on the uniform motion of the conveyor belt. It is difficult to use under the environment, and there is a defect that the equipment cost is high and cannot be used on a large scale

Images