CN114882499B - Fruit acid-sweetness classification method and system based on artificial intelligence - Google Patents
Fruit acid-sweetness classification method and system based on artificial intelligence Download PDFInfo
- Publication number
- CN114882499B CN114882499B CN202210797557.7A CN202210797557A CN114882499B CN 114882499 B CN114882499 B CN 114882499B CN 202210797557 A CN202210797557 A CN 202210797557A CN 114882499 B CN114882499 B CN 114882499B
- Authority
- CN
- China
- Prior art keywords
- fruit
- classification
- obtaining
- sweetness
- connected domain
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/60—Type of objects
- G06V20/68—Food, e.g. fruit or vegetables
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
- G06V10/267—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion by performing operations on regions, e.g. growing, shrinking or watersheds
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/764—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
Abstract
The invention discloses a fruit acid and sweetness classification method and system based on artificial intelligence, relates to the field of artificial intelligence, obtains appearance characteristics of fruits through image processing and classifies the fruit acid and sweetness, effectively improves classification efficiency and can avoid errors of manual operation. The method mainly comprises the following steps: acquiring a preset visual angle image of a fruit to be classified; carrying out image segmentation on the preset visual angle image to obtain a fruit segmentation image; obtaining a connected domain of the fruit by using a fruit segmentation graph, and establishing a plane rectangular coordinate system for the connected domain of the fruit; obtaining a distance curve according to the distance from the pixels at the edge in the connected domain to the origin of the plane rectangular coordinate system; obtaining the steepness of a distance curve; obtaining classification indexes of the fruits according to the steepness; and the classification of the acid sweetness of the fruits is realized by utilizing the classification indexes of the fruits. The specific application scenarios of the invention are as follows: in the fruit processing process, different varieties of the same fruit are classified through the appearance, and then the different types of acid-sweetness are utilized to realize the classification of the acid-sweetness.
Description
Technical Field
The application relates to the field of artificial intelligence, in particular to a fruit acid-sweetness classification method and system based on artificial intelligence.
Background
For the same fruit, different varieties often exist, the sour and sweet degrees of the different varieties are different, and the appearances of the different varieties are different; therefore, different varieties of the same fruit need to be separated, and the processing such as classified packaging, sale and the like is further facilitated.
At present, aiming at collected fruits, varieties of the same fruits are judged through the appearance of the fruits by utilizing naked eyes manually, and the sour-sweet degree classification is realized by utilizing the difference of the sour-sweet degree among different varieties.
Disclosure of Invention
The invention provides a fruit acid-sweetness classification method and system based on artificial intelligence, which comprises the following steps: acquiring a preset visual angle image of a fruit to be classified; carrying out image segmentation on the preset visual angle image to obtain a fruit segmentation image; obtaining a connected domain of the fruit by using a fruit segmentation graph, and establishing a plane rectangular coordinate system in the connected domain of the fruit; obtaining a distance curve according to the distance from the pixels at the edge in the connected domain to the origin of the plane rectangular coordinate system; obtaining the steepness of a distance curve; obtaining classification indexes of the fruits according to the steepness; the fruit sorting method has the advantages that fruit sorting is achieved by using fruit sorting indexes, compared with the prior art, the fruit sorting method has the advantages that fruit varieties are sorted by using the appearance characteristics of the fruits, the sour and sweet degrees of the fruits are sorted by using the difference of the sour and sweet degrees of different varieties of the same fruit, and subsequent sorting, packaging, selling and other treatment are facilitated; the appearance characteristics of the fruits are obtained through image processing, and the fruit varieties are classified by utilizing the appearance characteristics of the fruits, so that the classification efficiency is effectively improved, and meanwhile, subjective errors caused by artificial classification can be avoided.
Aiming at the technical problems, the invention provides a fruit acid-sweetness classification method and system based on artificial intelligence.
In a first aspect, provided herein is an artificial intelligence-based method for classifying fruit acid sweetness, comprising:
and acquiring a preset visual angle image of the fruit to be classified.
And carrying out image segmentation on the preset visual angle image to obtain a fruit segmentation image.
And obtaining a connected domain of the fruit by using the fruit segmentation graph, and establishing a plane rectangular coordinate system for the connected domain of the fruit.
And obtaining a distance curve according to the distance from the pixels at the edge in the connected domain to the origin of the plane rectangular coordinate system.
The steepness of the distance curve is obtained.
And obtaining the classification index of the fruit according to the steepness.
And classifying the acid sweetness of the fruit by using the classification index of the fruit.
Further, the artificial intelligence-based fruit sour-sweetness classification method for obtaining connected components of fruit by using the fruit segmentation map comprises the following steps:
and separating the first fruit part and the second fruit part from each other in the fruit segmentation image to obtain a connected domain of the fruit.
Further, the method for classifying fruit sour and sweet based on artificial intelligence, which uses the fruit segmentation map to obtain a connected domain of the fruit and establishes a rectangular plane coordinate system for the connected domain of the fruit, includes:
and carrying out open operation on the fruit segmentation graph to obtain a connected domain of a single fruit.
Obtaining a center of gravity of the connected domains of the individual fruit.
And taking the shortest connecting line of the two pixels on the edge of the connected domain of the single fruit passing through the central point as a horizontal axis.
And establishing a plane rectangular coordinate system by using the central point and the transverse axis.
Further, the artificial intelligence-based fruit sour sweetness classification method for obtaining the steepness of the distance curve comprises:
and dividing the distance curve into different distance curves according to the connected domain of the fruit in the quadrant of the plane rectangular coordinate system.
And acquiring the steepness of the different distance curves.
And obtaining the steepness of the distance curve according to the steepness of the distance curves of different sections.
Further, the fruit sour sweetness classification method based on artificial intelligence, which obtains the steepness of the distance curve, includes:
whereinIn order to be the steepness of the distance curve,is as followsThe vertical coordinate of the maximum point of the segment distance curve;is as followsThe abscissa of the maximum point of the segment distance curve;is as followsThe vertical coordinate of the minimum value point of the segment distance curve;is as followsThe maximum value of the abscissa of the segment distance curve;is as followsThe maximum value of the abscissa of the distance curve,。
further, the fruit acid-sweetness classification method based on artificial intelligence obtains a classification index of the fruit according to the steepness, and comprises the following steps:
the fruit is classified intoWhereinThe steepness of the distance curve of the fruit,is the aspect ratio of the fruitWhereinThe distance between two pixels intersected by the longitudinal axis of the rectangular plane coordinate system and the edge of the connected domain of the fruit isThe distance between two pixels of which the horizontal axis of the rectangular plane coordinate system is intersected with the edge of the connected domain of the fruit is obtained.
In a second aspect, the present invention provides an artificial intelligence-based fruit acid-sweetness classification system, including:
and the image acquisition module is used for acquiring the preset visual angle image of the fruit to be classified.
And the image segmentation module is used for carrying out image segmentation on the preset visual angle image to obtain a fruit segmentation image.
And the connected domain obtaining module is used for obtaining the connected domain of the fruit by using the fruit segmentation graph and establishing a plane rectangular coordinate system for the connected domain of the fruit.
And the distance curve acquisition module is used for acquiring a distance curve according to the distance from the pixels at the edge in the connected domain to the origin of the plane rectangular coordinate system.
And the abruptness calculation module is used for obtaining the abruptness of the distance curve.
And the classification index calculation module is used for obtaining the classification index of the fruit according to the steepness.
And the acid sweetness classification module is used for classifying the acid sweetness of the fruit by utilizing the classification index of the fruit.
The invention provides a fruit acid-sweetness classification method and system based on artificial intelligence, which comprises the following steps: acquiring a preset visual angle image of a fruit to be classified; carrying out image segmentation on the preset visual angle image to obtain a fruit segmentation image; obtaining a connected domain of the fruit by using a fruit segmentation graph, and establishing a plane rectangular coordinate system in the connected domain of the fruit; obtaining a distance curve according to the distance from the pixels at the edge in the connected domain to the origin of the plane rectangular coordinate system; obtaining the steepness of a distance curve; obtaining a classification index of the fruits according to the steepness; the fruit classification is realized by using the fruit classification indexes, compared with the prior art, the fruit classification method has the advantages that the fruit varieties are classified by using the appearance characteristics of the fruits, and the sour-sweet degree of the fruits is classified by using the difference of the sour-sweet degree between different varieties of the same fruit, so that the subsequent classification, packaging, sale and other treatment are facilitated; the appearance characteristics of the fruits are obtained through image processing, the fruit varieties are classified by utilizing the appearance characteristics of the fruits, the classification efficiency is effectively improved, and meanwhile subjective errors caused by artificial classification can be avoided.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to these drawings without creative efforts.
Fig. 1 is a schematic flow chart of an artificial intelligence-based fruit acid-sweetness classification method according to an embodiment of the present invention.
Fig. 2 is a schematic flow chart of another artificial intelligence-based fruit acid sweetness classification method provided by the embodiment of the invention.
Fig. 3 is a schematic diagram of a rectangular plane coordinate system of a connected component provided in an embodiment of the present invention.
Fig. 4 is a schematic diagram of a distance curve provided by an embodiment of the present invention.
Fig. 5 is a schematic flow chart of an artificial intelligence-based system for classifying the sweetness and acidity of a fruit according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
In the following description, for purposes of explanation and not limitation, specific details are set forth, such as particular system structures, techniques, etc. in order to provide a thorough understanding of the embodiments of the present application. It will be apparent, however, to one skilled in the art that the present application may be practiced in other embodiments that depart from these specific details. In other instances, detailed descriptions of well-known systems, devices, circuits, and methods are omitted so as not to obscure the description of the present application with unnecessary detail.
The terms "first", "second" and "first" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature; in the description of the present embodiment, "a plurality" means two or more unless otherwise specified.
Example 1
The embodiment of the invention provides an artificial intelligence-based fruit acid-sweetness classification method, which comprises the following steps of:
s101, acquiring a preset visual angle image of the fruit to be classified.
Wherein, the preset visual angle in this embodiment can be adjusted correspondingly according to the concrete type of current fruit, for example: if the current fruit is a fruit with a general fruit shape (apple, peach, persimmon, kiwi fruit and the like), selecting overlook as a preset visual angle; if the current fruit is the fruit with special fruit shape (carambola, banana and the like), the preset visual angle can be adjusted according to actual needs.
Since fruits are often transported on a conveyor belt in the production process, in the embodiment, a camera is used for shooting a preset visual angle image of the fruits to be classified, the image obtained in the embodiment is an RGB image, RGB is a color standard, and various colors are obtained by changing three color channels of red (R), green (G) and blue (B) and superposing the three color channels of red (R), green (G) and blue (B), wherein RGB is the color representing the three channels of red, green and blue.
And S102, carrying out image segmentation on the preset visual angle image to obtain a fruit segmentation image.
Image segmentation is a technique and process that divides an image into several specific regions with unique properties and proposes an object of interest. It is a key step from image processing to image analysis. The existing image segmentation methods mainly include the following categories: a threshold-based segmentation method, a region-based segmentation method, an edge-based segmentation method, a particular theory-based segmentation method, and the like. From a mathematical point of view, image segmentation is the process of dividing a digital image into mutually disjoint regions. The process of image segmentation is also a labeling process, i.e. pixels belonging to the same region are assigned the same number.
In this embodiment, a fruit part in an image is distinguished from a background part by image segmentation to obtain a fruit segmentation image, where the fruit segmentation image is a binary image, the binary image is an image in which each pixel is black or white, and the background part and the fruit part in the fruit segmentation image obtained in this embodiment are black and white.
S103, obtaining a connected domain of the fruit by using the fruit segmentation graph, and establishing a plane rectangular coordinate system for the connected domain of the fruit.
The connected component, also called connected component, refers to a pixel set composed of adjacent pixels with the same pixel value, in this embodiment, an opening operation is performed on the connected component in the fruit segmentation image, so as to separate a first fruit part and a second fruit part in the fruit segmentation image from each other, where the first fruit part and the second fruit part are connected components of the same fruit, and there is an overlapping portion between the first fruit part and the second fruit part, so as to obtain connected components of the fruit.
The opening operation is a method in image processing, which is used to eliminate small objects, objects separated at fine points, and smooth the boundary of a larger object without significantly changing its area.
And S104, obtaining a distance curve according to the distance from the pixels at the edge of the connected domain to the origin of the plane rectangular coordinate system.
Pixels are defined by tiles of the image that have a well-defined location and assigned color values that determine how the image appears.
In this embodiment, an intersection point of an edge of a connected domain and a positive half axis of a longitudinal axis of a planar rectangular coordinate system is taken as a starting point, the edge of the connected domain is traversed clockwise, an arc length accumulated in the traversing process is taken as an abscissa, a linear distance from the point to an origin of the planar rectangular coordinate system in the traversing process is taken as an ordinate, and a distance curve is obtained according to the abscissa and the ordinate, wherein traversing refers to sequentially making one visit to points in a searched route along a certain search route.
And S105, obtaining the steepness of the distance curve.
In this embodiment, the change of the distance curve is reflected by the steepness to obtain the change of the shape of the fruit, the distance curve is first divided into different segments of distance curves in a quadrant of a planar rectangular coordinate system according to the connected domain of the fruit, then the steepness of the different segments of distance curves is obtained, and finally the steepness of the distance curve is obtained according to the steepness of the different segments of distance curves.
And S106, obtaining a classification index of the fruit according to the steepness.
The aspect ratio of the fruit is taken into consideration in the embodiment, the aspect ratio in the embodiment refers to a ratio of the length of the fruit connected domain edge in the longitudinal axis of the planar rectangular coordinate system to the length of the fruit connected domain edge in the transverse axis of the planar rectangular coordinate system, the ratio of the height to the width of the fruit can be reflected through the aspect ratio, for example, when the fruit to be classified is kiwi fruit, different kiwi fruits may have a strip shape, a uniform shape of the length and the width, and the like, and finally, the classification index of the fruit is obtained by combining the steepness and a preset threshold value.
And S107, classifying the acid sweetness of the fruit by using the classification index of the fruit.
Generally, due to the continuous development of planting technology and agriculture, a variety of fruits exist, taking apples as an example, red fuji, akxose heart, galha, yellow marshal and the like exist in the existing market, the shapes of different varieties have great difference, and the acid sweetness of the tastes of different varieties have difference.
The classification of fruit sour sweetness degree is realized to the categorised index realization that utilizes the fruit that obtains in S106 in this embodiment, uses this kind of fruit of kiwi fruit as the example in this embodiment, learns through sour sweetness and appearance analysis to the different cultivars of kiwi fruit: for different varieties of kiwi fruits, the varieties with large classification indexes have large sweetness, namely small acidity, and the varieties with small classification indexes have small sweetness, namely large acidity.
Compared with the traditional technical scheme, the beneficial effects of the embodiment are as follows:
the appearance characteristics of the fruits are utilized to classify the varieties of the same fruit, and the sour-sweet degree of the fruits is classified by utilizing the difference of the sour-sweet degree between different varieties of the same fruit, so that the subsequent classified packaging, sale and other treatment are facilitated; the appearance characteristics of the fruits are obtained through image processing, and the fruit varieties are classified by utilizing the appearance characteristics of the fruits, so that the classification efficiency is effectively improved, and meanwhile, subjective errors caused by artificial classification can be avoided.
Example 2
The embodiment of the invention provides another fruit acid-sweetness classification method based on artificial intelligence, which is shown in fig. 2 and comprises the following steps:
s201, acquiring a preset visual angle image of the fruit to be classified.
Wherein, the preset visual angle in this embodiment can be adjusted correspondingly according to the concrete type of current fruit, for example: if the current fruit is a fruit with a general fruit shape (apple, peach, persimmon, kiwi fruit and the like), selecting overlook as a preset visual angle; if the current fruit is the fruit with special fruit shape (carambola, banana and the like), the preset visual angle can be adjusted according to actual needs.
Since fruits are often transported on a conveyor in the production process, in this embodiment, a camera is used to capture a preset viewing angle image of the fruits to be sorted on the conveyor, the image obtained in this embodiment is an RGB image, RGB is a color standard, and various colors are obtained by changing three color channels of red (R), green (G), and blue (B) and superimposing the three color channels on each other, wherein RGB is a color representing the three channels of red, green, and blue.
S202, carrying out image segmentation on the preset visual angle image to obtain a fruit segmentation image.
Image segmentation is a technique and process that divides an image into several specific regions with unique properties and proposes an object of interest. It is a key step from image processing to image analysis. The existing image segmentation methods mainly include the following categories: a threshold-based segmentation method, a region-based segmentation method, an edge-based segmentation method, a particular theory-based segmentation method, and the like. From a mathematical point of view, image segmentation is the process of dividing a digital image into mutually disjoint regions. The process of image segmentation is also a labeling process, i.e. pixels belonging to the same region are assigned the same number.
In this embodiment, a fruit part in the image is distinguished from a background part by image segmentation to obtain a fruit segmentation image, where the fruit segmentation image is a binary image, the binary image is an image in which each pixel is black or white, and the background part and the fruit part in the fruit segmentation image obtained in this embodiment are black and white.
In this embodiment, image segmentation is performed by adopting a semantic segmentation method of Deep Neural Network (DNN), and the content of the DNN is as follows:
(1) the data set used is a fruit image data set on the conveyor belt acquired from above.
(2) The pixels needing to be segmented are divided into 2 types, namely the labeling process of the corresponding labels of the training set is as follows: pixels belonging to the background are labeled 0 and pixels belonging to the fruit are labeled 1.
(3) The loss function used is cross entropy.
After the network training is completed in this embodiment, the preset view angle image of the fruit to be classified may be input into the DNN to obtain a fruit segmentation image.
S203, obtaining the connected domain of the fruit by using the fruit segmentation graph, and establishing a plane rectangular coordinate system for the connected domain of the fruit.
In this embodiment, S203 specifically includes S2031, S2032, S2033, and S2034.
S2031, performing open operation on the fruit segmentation graph to obtain a connected domain of a single fruit.
Since the multiple fruits are contacted with each other and can be identified as the same connected domain, the connected domain in the fruit segmentation image is subjected to the opening operation in the embodiment, the first fruit part and the second fruit part in the fruit segmentation image are separated from each other, the first fruit part and the second fruit part have an overlapping part, the first fruit part and the second fruit part are the connected domain of the same fruit, and therefore the connected domain of the fruit is obtained, and only one fruit exists in the connected domain of each fruit in the embodiment.
The opening operation is a method in image processing, which is used to eliminate small objects, objects separated at fine points, and smooth the boundary of a larger object without significantly changing the area of the object.
The connected component is also referred to as a connected component, and refers to a set of pixels composed of adjacent pixels having the same pixel value.
S2032, obtaining the central point of the connected domain of the single fruit through the connected domain analysis。
S2033, passing two pixels on the edge of the connected domain of the single fruit through the center pointThe shortest connecting line of (2) is taken as the horizontal axis.
Specifically, the over-center point is calculated in this embodimentObtaining the length set of all line segments by the line segments intersected with the edges of the connected domainWherein the firstHas a length ofGet a setThe shortest line segment in the set is taken as the horizontal axis of the connected domain.
S2034, a plane rectangular coordinate system is established by the central point and the transverse axis.
Taking a line segment which is perpendicular to the transverse axis and crossed with the edge point of the connected domain by the over-center point as the longitudinal axis of the connected domain, and taking the center point of the connected domainAnd establishing a plane rectangular coordinate system by taking the original point as the horizontal axis as the x axis and the vertical axis as the y axis. FIG. 3 is a schematic diagram of a rectangular coordinate system of the plane in the present embodiment, as shown in FIG. 3, the center pointHas the coordinates ofThe coordinates of two end points of the cross axis and the edge of the connected domain are respectively 、 Wherein. If it isThe connected components are then processed symmetrically about the vertical, y-axis.
The coordinates of two end points of which the longitudinal axis intersects with the edge of the connected domain are respectively 、 The distance from the intersection point of the positive half axis of the longitudinal axis and the edge of the connected domain to the center point isThe distance between the negative half axis of the longitudinal axis and the edge of the connected domain from the center point isWhereinIf, ifThe connected component is processed symmetrically about the horizontal axis, i.e., the x-axis.
And S204, obtaining a distance curve according to the distance from the pixels at the edge of the connected domain to the origin of the plane rectangular coordinate system.
In this embodiment, an intersection point of an edge of the connected domain and a positive half axis of a longitudinal axis of the planar rectangular coordinate system is taken as a starting point, the edge of the connected domain is traversed clockwise, traversal is ended when the edge returns to the starting point, an arc length accumulated in the traversal process is taken as an abscissa, and a second arc length accumulated in the traversal process is taken as an abscissaDotLinear distance to origin in a rectangular plane coordinate systemIs the ordinate in whichAnd obtaining a distance curve according to the abscissa and the ordinate, wherein traversing means that all points in a searched route are visited once in sequence along a certain searched route.
As shown in fig. 4, of the abscissa of a point in the distance curveSequentially corresponding to connected domain edge pointsThe cumulative arc length of the edges is initially traversed in a clockwise direction,for the length of the arc over which the traversal is started, i.e.=0;Rectangular plane coordinate system of connected domainClockwise along the edge of the connected domain toThe length of the arc of (a) is,in a rectangular plane coordinate system of connected domainsClockwise along the edge of the connected domain toThe length of the arc of (a) is,in a rectangular plane coordinate system of connected domainsClockwise along the edge of the connected domain toThe length of the arc of (a) is,the length of the accumulated arc line is the perimeter of the edge of the connected domain when the edge of the whole connected domain is traversed clockwise; the ordinate of the midpoint of the distance curve represents the distance from the edge point in the connected domain to the center point of the connected domain in performing the clockwise traversal.
And S205, obtaining the steepness of the distance curve.
The change degree of the edge arc of the fruit in different quadrants is reflected by each stage of the distance curve, and the change condition of the edge arc of each quadrant of the fruit is obtained by analyzing the steepness of each stage of the distance curve, so that the change condition of the whole edge arc of the fruit is obtained.
In this embodiment, S205 specifically includes S2051, S2052, and S2053.
S2051, dividing the distance curve into different distance curves according to connected domains of the fruits in a quadrant of the plane rectangular coordinate system.
Since the abscissa of the midpoint of the distance curve representsStart of sequenceThe hour hand traverses the accumulated arc length of the edge point of the connected domain of the fruit, so the distance curve is divided into four segments according to the quadrant in the plane rectangular coordinate system of the connected domain in the embodiment 、 、 、 The first stage isThe second stage isThe third stage isThe fourth section is。
And S2052, acquiring the steepness of the distance curves of different segments.
Wherein the content of the first and second substances,is a distance curve inThe ordinate of the maximum point of each stage indicates that the fruit is in the second stageThe maximum distance from the edge points to the center point in the quadrant,is a distance curve inThe abscissa of the maximum point of each phase.Is a distance curve inThe ordinate of the minimum point of each phase.
Analyzing the steepness of each distance curve, wherein the greater the steepness of a certain distance curve is, the greater the change of the distance curve in the section is, and the more curved the edge arc line of the fruit image in the corresponding quadrant is; the less steep a certain distance curve is, the more gradual the distance curve is at that stage, and the more gradual the edge arc is in the corresponding quadrant for the fruit image.
And S2053, acquiring the steepness of the distance curve according to the steepness of the distance curves of different segments.
Reflects the distance curve inThe change of the stage, to measure the overall change of the distance curve, isWeighted average is calculated to obtain the steepness of the whole distance curveWhereinIs as followsThe weight coefficient of the stage reflects the ratio of the arc length of the corresponding fruit in the total circumference, and
and S206, obtaining the classification index of the fruit according to the steepness.
First, the aspect ratio of the connected region of the single fruit is determined,WhereinThe distance between two pixels intersected by the longitudinal axis of the rectangular plane coordinate system and the edge of the connected domain of the fruit isThe distance between two pixels where the horizontal axis of the rectangular plane coordinate system intersects with the edge of the connected domain of the fruit is shown in this embodiment。
Then defining the classification index asIn this embodiment, the classification indexes are divided into three types, and the implementer can determine the total number of the classified varieties according to specific requirements, and the first classification threshold and the second classification threshold are combined to obtain the classification indexesThe first classification threshold in this embodiment isThe second classification threshold isIn this embodiment, the specific SD value calculation method is as follows:
wherein]To round up, the data is, illustratively,when the temperature of the water is higher than the set temperature,i.e. SD =1 at this time.
And S207, classifying the acid sweetness of the fruits by using the classification indexes of the fruits.
Generally, due to the continuous development of planting technology and agriculture, a variety of fruits exist, taking apples as an example, red fuji, akxose heart, galha, yellow marshal and the like exist in the existing market, the shapes of different varieties have great difference, and the acid sweetness of the tastes of different varieties have difference.
Utilize the categorised index of fruit that obtains in S206 to realize the classification to fruit sour sweetness in this embodiment, use this kind of fruit of kiwi fruit as the example in this embodiment, learn through the sour sweetness and the appearance to the different cultivars of kiwi fruit to analyze: for different varieties of kiwi fruits, the varieties with large classification indexes have large sweetness, namely small acidity, and the varieties with small classification indexes have small sweetness, namely large acidity; therefore, in this embodiment, the kiwi fruit variety with the classification index SD =1 is sour, the kiwi fruit variety with the classification index SD =2 is moderate in sour and sweet, and the kiwi fruit variety with the classification index SD =3 is sweet.
Compared with the traditional technical scheme, the beneficial effects of the embodiment are as follows:
the appearance characteristics of the fruits are utilized to classify the varieties of the same fruit, and the sour-sweet degree of the fruits is classified by utilizing the difference of the sour-sweet degree between different varieties of the same fruit, so that the subsequent classified packaging, sale and other treatment are facilitated; the appearance characteristics of the fruits are obtained through image processing, and the fruit varieties are classified by utilizing the appearance characteristics of the fruits, so that the classification efficiency is effectively improved, and meanwhile, subjective errors caused by artificial classification can be avoided.
Example 3
The embodiment of the invention provides an artificial intelligence-based fruit acid-sweetness classification system, which is characterized by comprising the following components in percentage by weight as shown in figure 5:
the image obtaining module 301 is configured to obtain a preset view image of a fruit to be classified.
The image segmentation module 302 is configured to perform image segmentation on the preset view image to obtain a fruit segmentation image.
And the connected domain obtaining module 303 is configured to obtain a connected domain of the fruit by using the fruit segmentation map, and establish a rectangular plane coordinate system for the connected domain of the fruit.
A distance curve obtaining module 304, configured to obtain a distance curve according to a distance from a pixel at an edge in the connected domain to an origin of the planar rectangular coordinate system.
A steepness calculation module 305 for obtaining a steepness of the distance curve.
And a classification index calculation module 306, configured to obtain a classification index of the fruit according to the steepness.
And the sour-sweet degree classification module 307 is configured to classify the sour-sweet degree of the fruit by using the classification index of the fruit.
Here, it should be noted that: for detailed description of each module in this embodiment, reference may be made to other embodiments, and details are not repeated herein.
In conclusion, compared with the prior art, the fruit type classification method has the advantages that the appearance characteristics of the fruits are utilized to classify the varieties of the same fruit, and the sour-sweet degree of the fruits is classified by utilizing the difference of the sour-sweet degree of different varieties of the same fruit, so that the subsequent classification, packaging, sale and other treatment are facilitated; the appearance characteristics of the fruits are obtained through image processing, and the fruit varieties are classified by utilizing the appearance characteristics of the fruits, so that the classification efficiency is effectively improved, and meanwhile, subjective errors caused by artificial classification can be avoided.
The use of words such as "including," "comprising," "having," and the like, in the present invention is an open-ended word that refers to "including, but not limited to," and that may be used interchangeably therewith. The words "or" and "as used herein mean, and are used interchangeably with, the word" and/or, "unless the context clearly dictates otherwise. The word "such as" is used herein to mean, and is used interchangeably with, the phrase "such as but not limited to".
It should also be noted that the various components or steps may be broken down and/or re-combined in the methods and systems of the present invention. These decompositions and/or recombinations are to be considered equivalents of the present disclosure.
The above-mentioned embodiments are merely examples for clearly illustrating the present invention and do not limit the scope of the present invention. It will be apparent to those skilled in the art that other variations and modifications may be made in the foregoing description, and it is not necessary or necessary to exhaustively enumerate all embodiments herein. All designs identical or similar to the present invention are within the scope of the present invention.
Claims (6)
1. An artificial intelligence-based fruit acid-sweetness classification method is characterized by comprising the following steps:
acquiring a preset visual angle image of a fruit to be classified;
carrying out image segmentation on the preset visual angle image to obtain a fruit segmentation image;
obtaining a connected domain of the fruit by using the fruit segmentation graph, and establishing a plane rectangular coordinate system for the connected domain of the fruit;
obtaining a distance curve according to the distance from the pixels at the edge of the connected domain to the origin of the plane rectangular coordinate system;
obtaining the steepness of the distance curve;
obtaining a classification index of the fruit according to the steepness;
the classification of the acid sweetness of the fruit is realized by utilizing the classification index of the fruit;
wherein, obtaining a classification index of the fruit according to the steepness comprises:
whereinIs a classification index of fruit [ alpha ], [ beta ], [ alpha ], [ beta ], [ alpha ], [ beta ] or a]In order to get the whole upwards,is the abruptness of the distance curve of the fruit,is the aspect ratio of the fruitIn whichThe distance between two pixels intersected by the longitudinal axis of the rectangular plane coordinate system and the edge of the connected domain of the fruit isThe distance between two pixels of which the horizontal axis of the rectangular plane coordinate system intersects with the edge of the connected domain of the fruit,is a natural constantCounting;
wherein, utilize the categorised index of fruit realizes the classification to fruit acid sweetness, includes: the kiwi fruit variety with the classification index SD =1 is sour, the kiwi fruit variety with the classification index SD =2 is moderate in sweetness and sourness, and the kiwi fruit variety with the classification index SD =3 is sweet.
2. The method as claimed in claim 1, wherein the step of obtaining the connected components of the fruit by using the fruit segmentation graph comprises:
and separating the first fruit part and the second fruit part from each other in the fruit segmentation image to obtain a connected domain of a single fruit.
3. The artificial intelligence based fruit sour sweetness classification method according to claim 1, wherein the obtaining of the connected region of the fruit by using the fruit segmentation map and the establishing of the plane rectangular coordinate system for the connected region of the fruit comprises:
performing opening operation on the fruit segmentation graph to obtain a connected domain of a single fruit;
obtaining a center point of the single fruit through connected domain analysis;
taking the shortest connecting line of two pixels on the edge of the connected domain of the single fruit passing through the central point as a horizontal axis;
and establishing a plane rectangular coordinate system by using the central point and the transverse axis.
4. The artificial intelligence based fruit sour sweetness classification method of claim 1, wherein obtaining the steepness of the distance curve comprises:
dividing the distance curve into different sections of distance curves in a quadrant of the plane rectangular coordinate system according to the connected domain of the fruit;
acquiring the steepness of the different distance curves;
and obtaining the steepness of the distance curve according to the steepness of the distance curves of different sections.
5. The artificial intelligence based fruit sour sweetness classification method of claim 4, wherein obtaining the steepness of the distance curve comprises:
wherein the content of the first and second substances,in order to be the steepness of the distance curve,is as followsThe ordinate of the maximum point of the segment distance curve,is as followsThe abscissa of the maximum point of the segment distance curve,is as followsThe ordinate of the minimum point of the segment distance curve,is as followsThe maximum value of the abscissa of the segment distance curve,is as followsThe maximum value of the abscissa of the distance curve,。
6. an artificial intelligence based fruit acid and sweetness classification system, comprising:
the image acquisition module is used for acquiring a preset visual angle image of the fruit to be classified;
the image segmentation module is used for carrying out image segmentation on the preset visual angle image to obtain a fruit segmentation image;
the connected region acquisition module is used for acquiring the connected regions of the fruits by using the fruit segmentation graph and establishing a plane rectangular coordinate system for the connected regions of the fruits;
the distance curve acquisition module is used for acquiring a distance curve according to the distance from the pixels at the edge in the connected domain to the origin of the plane rectangular coordinate system;
a steepness calculation module for obtaining a steepness of the distance curve;
the classification index calculation module is used for obtaining the classification index of the fruit according to the steepness;
the acid-sweetness classification module is used for classifying the acid sweetness of the fruit by using the classification index of the fruit;
wherein, obtaining a classification index of the fruit according to the steepness comprises:
whereinIs a classification index of fruit [ alpha ], [ beta ], [ alpha ], [ beta ], [ alpha ], [ beta ] or a]In order to get the whole upwards,the steepness of the distance curve of the fruit,is the aspect ratio of the fruitWhereinThe distance between two pixels intersected by the longitudinal axis of the rectangular plane coordinate system and the edge of the connected domain of the fruit isThe distance between two pixels of which the horizontal axis of the rectangular plane coordinate system intersects with the edge of the connected domain of the fruit,is a natural constant;
wherein, utilize the categorised index of fruit realizes the classification to fruit acid sweetness, includes: the kiwi fruit variety with the classification index SD =1 is sour, the kiwi fruit variety with the classification index SD =2 is moderate in sour and sweet, and the kiwi fruit variety with the classification index SD =3 is sweet.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210797557.7A CN114882499B (en) | 2022-07-08 | 2022-07-08 | Fruit acid-sweetness classification method and system based on artificial intelligence |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210797557.7A CN114882499B (en) | 2022-07-08 | 2022-07-08 | Fruit acid-sweetness classification method and system based on artificial intelligence |
Publications (2)
Publication Number | Publication Date |
---|---|
CN114882499A CN114882499A (en) | 2022-08-09 |
CN114882499B true CN114882499B (en) | 2022-09-20 |
Family
ID=82682628
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202210797557.7A Active CN114882499B (en) | 2022-07-08 | 2022-07-08 | Fruit acid-sweetness classification method and system based on artificial intelligence |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN114882499B (en) |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2007195024A (en) * | 2006-01-20 | 2007-08-02 | Sony Corp | Image processing apparatus and method, learning apparatus and method, and program |
CN109115775A (en) * | 2018-08-08 | 2019-01-01 | 长沙理工大学 | A kind of betel nut level detection method based on machine vision |
CN111968144A (en) * | 2020-09-07 | 2020-11-20 | 北京凌云光技术集团有限责任公司 | Image edge point acquisition method and device |
-
2022
- 2022-07-08 CN CN202210797557.7A patent/CN114882499B/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2007195024A (en) * | 2006-01-20 | 2007-08-02 | Sony Corp | Image processing apparatus and method, learning apparatus and method, and program |
CN109115775A (en) * | 2018-08-08 | 2019-01-01 | 长沙理工大学 | A kind of betel nut level detection method based on machine vision |
CN111968144A (en) * | 2020-09-07 | 2020-11-20 | 北京凌云光技术集团有限责任公司 | Image edge point acquisition method and device |
Also Published As
Publication number | Publication date |
---|---|
CN114882499A (en) | 2022-08-09 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US9852344B2 (en) | Systems and methods for semantically classifying and normalizing shots in video | |
CN107392925B (en) | Remote sensing image ground object classification method based on super-pixel coding and convolutional neural network | |
Jidong et al. | Recognition of apple fruit in natural environment | |
US8311344B2 (en) | Systems and methods for semantically classifying shots in video | |
US9020263B2 (en) | Systems and methods for semantically classifying and extracting shots in video | |
CN111428781A (en) | Remote sensing image ground object classification method and system | |
CN109636784B (en) | Image saliency target detection method based on maximum neighborhood and super-pixel segmentation | |
CN110310241B (en) | Method for defogging traffic image with large air-light value by fusing depth region segmentation | |
CN110866879B (en) | Image rain removing method based on multi-density rain print perception | |
CN104834933A (en) | Method and device for detecting salient region of image | |
CN107622280B (en) | Modularized processing mode image saliency detection method based on scene classification | |
JP2010511215A (en) | How to indicate an object in an image | |
CN110717921B (en) | Full convolution neural network semantic segmentation method of improved coding and decoding structure | |
CN109871900A (en) | The recognition positioning method of apple under a kind of complex background based on image procossing | |
CN110728640A (en) | Double-channel single-image fine rain removing method | |
CN111291818B (en) | Non-uniform class sample equalization method for cloud mask | |
CN108596244A (en) | A kind of high spectrum image label noise detecting method based on spectrum angle density peaks | |
CN114882499B (en) | Fruit acid-sweetness classification method and system based on artificial intelligence | |
Zhao et al. | Image dehazing based on haze degree classification | |
CN110033443B (en) | Display panel defect detection method | |
CN114550167B (en) | Artificial intelligence based pear quality classification method and device | |
Nawawi et al. | Comprehensive pineapple segmentation techniques with intelligent convolutional neural network | |
CN115063603B (en) | Wood annual ring line accurate extraction and restoration method based on edge information | |
CN115330721A (en) | Banana fruit comb plumpness detection method and system based on shape and color information | |
CN115049865A (en) | Tobacco leaf grading method and system based on computer vision |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |