An Image Quantitative Detection Method for Carrot Defects

Abstract

The invention discloses a carrot defect image quantitative detection method and a carrot sorting apparatus. The quantitative detection method is used for detecting carrot defects like green heads, fibrous roots and cracking and comprises the following steps: collecting images of cleaned carrots with an industrial camera, transmitting the images to a computer for processing and carrying out binaryzation and edge detection by using an H component; then respectively carrying out detection of green heads, fibrous roots and cracking, wherein during detection of green heads, detection and quantification of green heads and cracking are carried out through an R component and an S component, and detection of fibrous roots is carried out by extracting skeletons and calculating the number of end points; and controlling the sorting apparatus to carry out sorting by using the computer according to detection results. With the method, quantitative detection of green heads, fibrous roots and cracking of carrots and sorting of carrots are realized, subjectivity of artificial detection is overcome, quantitative detection is more objective and scientific, and the method improves production efficiency when applied to agricultural production, quality grading and commerce circulation.

Description

technical field [0001] The invention relates to a method for detecting appearance quality defects of agricultural products used in the field of agricultural scientific research, in particular to an image quantitative detection method for green heads, fibrous roots and cracks of carrots. Background technique [0002] The grading and sales of carrots help to improve the market competitiveness of carrots and increase economic benefits. At present, some carrot production and processing enterprises mainly rely on artificial naked eyes to do some simple grading. However, these methods of detection and grading increase manpower and increase production costs. , and the production efficiency is low, the profit cannot be greatly improved, and it is not suitable for large-scale production and promotion. Some fruit and vegetable production and processing enterprises use the mechanical method of using rollers with different gaps according to the thickness of carrots instead of manual carr...

AI Technical Summary

Problems solved by technology

Although there are literature reports on computer image processing methods for agricultural products such as apples, citrus, and potatoes, these are basically concentrated on spherical and spherical fruits and vegetables. Due to the inherent limitations of the method, non-spherical fruits and vegetables such as carrots Apps are limited

Chinese patent application 201010210313.1 discloses a new type of carrot cleaning and sorting machine, but it is clear in this patent application that the sorting of carrot share and quality is achieved by manual sorting at the manual sorting station, and does not involve effective image processing. automatic sorting method

Images