CN109934817B

CN109934817B - Method for detecting malformation of external contour of fruit body

Info

Publication number: CN109934817B
Application number: CN201910217215.1A
Authority: CN
Inventors: 裴悦琨; 叶家敏; 连明月; 姜艳超
Original assignee: Dalian University
Current assignee: Dalian University
Priority date: 2019-03-21
Filing date: 2019-03-21
Publication date: 2023-02-03
Anticipated expiration: 2039-03-21
Also published as: CN109934817A

Abstract

The invention discloses a method for detecting the malformation of the external contour of a fruit body, which comprises the following steps: s1, collecting a color image of the external outline of a fruit, and converting the color image into a gray image; s2, carrying out image segmentation processing on the gray-scale image, and extracting the boundary area of the fruit body in the image by adopting a canny algorithm; s3, adopting a pixel tracking method to clockwise obtain pixel point coordinates of the fruit body boundary; s4, judging the concavity and convexity of the external contour of the fruit body by adopting a right-hand rule; and S5, comparing the number of concave particles on the external contour of the fruit body with a preset threshold value to judge whether the fruit body is removed or reserved. The method is based on computer software to detect the concave-convex defect degree of the fruit body surface, so that the malformed fruit body is removed at a certain precision and speed, a good foundation is laid for the subsequent grading, and the grading quality of the fruit body is improved.

Description

Method for detecting external contour deformity of fruit body

Technical Field

The invention relates to the technical field of fruit body external defect detection, in particular to a fruit body external contour malformation detection method.

Background

In the fruit malformation detection technology, the blade moves on the elevator through the blade opening on the hardware in the prior art, and the abnormal part of the fruit can be cut off by the blade under the action of pressing the fruit downwards. In addition, in the prior art, the defect detection method for the fruit in the can body utilizes the light axis to irradiate along the surface tangential direction of the can body, so that the irregular reflection and shadow are generated at the concave-convex part, and the defect part of the fruit body is detected. However, fruit defect detection methods in the prior art are based on industrial mechanical devices to detect fruit defects, so that not only is technical support of hardware equipment required, but also the detection precision is low due to external interference in the fruit defect detection process, and even after the fruit defects are detected, malformed fruits still appear.

Disclosure of Invention

According to the problems in the prior art, the invention discloses a method for detecting the malformation of the external contour of a fruit body, which specifically comprises the following steps:

s1, collecting a color image of the external outline of a fruit, and converting the color image into a gray image;

s2, carrying out image segmentation processing on the gray level image, removing fruit stalks by utilizing morphological open operation, carrying out denoising processing on fruit bodies by utilizing morphological close operation, and extracting the boundary area of the fruit bodies in the image by adopting a canny algorithm;

s3, adopting a pixel tracking method to clockwise obtain the pixel point coordinates of the fruit body boundary;

s4, judging the concave-convex property of the external outline of the fruit body by adopting a right-hand rule;

and S5, comparing the number of the concave particles on the external contour of the fruit body with a preset threshold value to judge whether the fruit body is removed or reserved.

Further, the concavity and convexity of the outer contour are specifically as follows:

selecting three pixel points of the fruit body boundary, taking m as a base point, setting m + lambda as the pixel point of the mth pixel point which is shifted forward by lambda, setting m-lambda as the pixel point of the mth pixel point which is shifted backward by lambda, and carrying out vector quantity

Is a vector formed by base points and backward shifted lambda-bit pixel points

Is a vector formed by base points and forward shifted lambda pixel points, and theta is a vector

Sum vector

Angle of formation, will vector

Sum vector

And (3) judging the orientation of a normal vector vertical plane by adopting right-hand determination for cross multiplication, judging that the external outline of the fruit body is convex if the normal vector vertical plane faces outwards, judging that the external outline of the fruit body is concave if the normal vector vertical plane faces inwards, moving m clockwise backwards by one bit, and sequentially and circularly judging the concavity and convexity of the boundary pixel points of the fruit body.

Further, the fruit body is removed or retained by adopting the following specific method:

calculating the number x of concave particles, comparing the number x with a threshold th, and when x is larger than or equal to th, determining that the fruit body is a defect and rejecting the fruit body; when x is less than th, the fruit body is regarded as normal and reserved;

due to the technical scheme, the method for detecting the malformation of the external contour of the fruit body is based on the software used by a computer to detect the concave-convex defect degree of the surface of the fruit body, so that the malformed fruit body is removed at a certain precision and speed, a good basis is laid for the subsequent grading, and the grading quality of the fruit body is improved.

Drawings

In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings needed to be used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments described in the present application, and other drawings can be obtained by those skilled in the art without creative efforts.

FIG. 1 is a flow chart of the method of the present invention;

FIG. 2 is a schematic diagram of an embodiment of the present invention;

FIG. 3 is a schematic diagram of an embodiment of the present invention.

Detailed Description

In order to make the technical solutions and advantages of the present invention clearer, the following describes the technical solutions in the embodiments of the present invention clearly and completely with reference to the drawings in the embodiments of the present invention:

as shown in fig. 1, the method for detecting the malformation of the external contour of a fruit body specifically comprises the following steps:

1. the RGB color mode is used to collect image information outside the fruit body, but RGB does not actually reflect morphological features of the image, so the color image needs to be grayed and converted into an 8-bit grayscale image, and then processed, for example, histogram, grayscale change, binarization and other operations are performed.

2. Dividing the gray image into image, and removing cherry fruit stem and shape by using morphological opening operation

3. And (4) denoising the fruit body by the closed-loop operation, and finally extracting the boundary of the fruit body by adopting a canny algorithm. And acquiring the coordinates of the pixel points of the fruit body boundary clockwise by adopting an 8-connected pixel tracking method on the denoised fruit body boundary.

4. As shown in fig. 2, three pixel points of the boundary of the fruit body are taken, m is taken as a base point, m can rotate clockwise, λ is the step length of the adjacent pixel point, m + λ is the pixel point of the mth pixel point which is shifted forward by λ, and m- λ is the pixel point of the mth pixel point which is shifted backward by λ, and a vector is obtained

Is a vector formed by base points and backward shifted lambda-bit pixel points

Is a vector formed by a base point and a forward-moving lambda-bit pixel point, theta is an included angle formed by the two vectors, and the right polygon approximately replaces the normal cherry fruit body. For convex edge knowledge, vector

Sum vector

The right-handed rule for cross-multiplication can determine that the normal vector vertical plane is outward. As shown in fig. 2 and 3, the right polygon is approximately substituted for the body of the cherry with the defect, and the concave edge is known as the vector

Sum vector

The normal vector obtained by cross multiplication faces inward in the vertical plane. And m, moving clockwise one bit backward, and sequentially and circularly judging the concavity and convexity of the boundary pixel points of the fruit body.

5. Calculating the number of concave particles and comparing the number with a preset threshold value, wherein x is the number of the concave particles and th is the preset threshold value as shown in formula (1), when x is larger than or equal to th, more points with deep concave degree on the fruit body are obtained, and the fruit body is regarded as a defect and is removed; and when x is less than th, the points with deep sunken degrees on the fruit body are less, and the fruit body is regarded as normal and is reserved for the next detection.

The above description is only for the preferred embodiment of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art should be considered to be within the technical scope of the present invention, and the technical solutions and the inventive concepts thereof according to the present invention should be equivalent or changed within the scope of the present invention.

Claims

1. A method for detecting the outer contour deformity of a fruit body is characterized by comprising the following steps:

s2, carrying out image segmentation processing on the gray image, removing fruit stalks by using morphological open operation, carrying out denoising processing on fruit bodies by using morphological close operation, and extracting the boundary region of the fruit bodies in the image by using a canny algorithm;

s4, judging the concavity and convexity of the external contour of the fruit body by adopting a right-hand rule;

s5, comparing the number of concave particles on the external contour of the fruit body with a preset threshold value to judge whether the fruit body is removed or reserved;

the method for judging the concave-convex of the external contour of the fruit body specifically adopts the following mode:

Is a vector formed by base points and backward shifted lambda-bit pixel points

Is a vector formed by base points and advanced lambda-bit pixel points, and theta is a vector

Sum vector

Form an angle of separation, a vector

Sum vector

Performing cross multiplication and judging the orientation of the normal vector vertical plane by adopting right-hand determination, and if the normal vector vertical plane faces outwards, judgingAnd (3) cutting the external contour of the fruit body into a convex edge, judging that the external contour of the fruit body is a concave edge if the normal vector vertical plane faces inwards, moving m clockwise backwards for one bit, and sequentially and circularly judging the concavity and convexity of the boundary pixel points of the fruit body.

2. The method for detecting malformation of exterior contour of fruit body according to claim 1, further characterized by: the specific method for judging whether the fruit body is removed or reserved is as follows: