CN110929787A - Apple objective grading system based on images - Google Patents
Apple objective grading system based on images Download PDFInfo
- Publication number
- CN110929787A CN110929787A CN201911158969.0A CN201911158969A CN110929787A CN 110929787 A CN110929787 A CN 110929787A CN 201911158969 A CN201911158969 A CN 201911158969A CN 110929787 A CN110929787 A CN 110929787A
- Authority
- CN
- China
- Prior art keywords
- image
- fruit
- apple
- camera
- grading
- 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.)
- Granted
Links
- 235000013399 edible fruits Nutrition 0.000 claims abstract description 55
- 238000000034 method Methods 0.000 claims abstract description 28
- 201000010099 disease Diseases 0.000 claims abstract description 24
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 claims abstract description 24
- 238000000605 extraction Methods 0.000 claims abstract description 11
- 230000007547 defect Effects 0.000 claims description 15
- 239000011159 matrix material Substances 0.000 claims description 10
- 238000012545 processing Methods 0.000 claims description 8
- 238000012847 principal component analysis method Methods 0.000 claims description 5
- 239000003086 colorant Substances 0.000 claims description 4
- 230000000877 morphologic effect Effects 0.000 claims description 3
- 238000006243 chemical reaction Methods 0.000 claims description 2
- 238000013145 classification model Methods 0.000 claims description 2
- 230000011218 segmentation Effects 0.000 claims description 2
- 241000220225 Malus Species 0.000 abstract description 47
- 235000021016 apples Nutrition 0.000 abstract description 20
- 238000001514 detection method Methods 0.000 abstract description 5
- 239000013598 vector Substances 0.000 description 8
- 238000005516 engineering process Methods 0.000 description 6
- 241000238631 Hexapoda Species 0.000 description 5
- 238000012549 training Methods 0.000 description 3
- 238000011161 development Methods 0.000 description 2
- 230000018109 developmental process Effects 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 241001270131 Agaricus moelleri Species 0.000 description 1
- 235000014443 Pyrus communis Nutrition 0.000 description 1
- 230000006835 compression Effects 0.000 description 1
- 238000007906 compression Methods 0.000 description 1
- 230000002950 deficient Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000000053 physical method Methods 0.000 description 1
- 230000001737 promoting effect Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
- G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/56—Extraction of image or video features relating to colour
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Multimedia (AREA)
- Probability & Statistics with Applications (AREA)
- Image Analysis (AREA)
Abstract
The invention provides an image-based apple objective grading method, which comprises the following steps: extracting disease characteristics of the single fruit image shot by the camera to judge whether the single fruit belongs to inedible categories; and performing classification feature extraction on the single fruits of the edible category, and grading the single fruits according to multi-feature indexes generated by feature extraction results. The invention replaces manpower with computer vision, replaces a large amount of labor force and avoids subjectivity and fatigue of the manpower; before grading work, inedible apples are eliminated, and a relatively complete system is realized; and the quality and the efficiency are improved compared with the traditional single characteristic by adopting multi-characteristic indexes for detection and classification.
Description
Technical Field
The invention relates to the technical field of image recognition and classification, in particular to an apple objective grading system based on images.
Background
With the development of high and new technologies, the combination of the high and new technologies and the agricultural field is deepened more and more, and the method plays an important role in promoting the development of agriculture. Machine vision technology was proposed in the 60's of the 20 th century as a representative of high-tech technology, and related research was started in China by the 80's. In recent years, machine vision technology has achieved abundant results in the field of fine agriculture, but China is a world fruit tree big country, the cultivation history is long, resources are abundant, fruits and dry fruits can reach more than 50 kinds, and the fruit tree is one of the countries with the earliest origin and the most kinds of fruit trees in the world.
The apples are the fruits with the largest yield in China, the export quantity is large, but the export price is low, and the accurate classification of the apple quality is an important link in the commercialized treatment of the apples mainly because the detection and classification technology of the picked apples in China lags behind and the competitiveness of the international market is lacked. Traditional fruit is hierarchical relies on manual operation and judges, and hierarchical speed is low, the precision is low, has great subjectivity, and degree of automation is low, and the human cost also begins to rise, can not satisfy international requirement. Therefore, how to rapidly, objectively and accurately grade apples becomes the research focus in the field of fruit grading at present.
Disclosure of Invention
According to the technical problems of low fruit grading efficiency, poor precision and serious missed detection phenomenon, the objective apple grading system based on the image is provided, the defect types of unqualified apples can be detected before grading, the unqualified apples are eliminated, and then automatic grading of the apples is carried out according to the 'quality grade requirement of fresh apples' issued by the state and various external characteristics of the apples.
The technical means adopted by the invention are as follows:
an image-based apple objective grading method is characterized by comprising the following steps: extracting disease characteristics of the single fruit image shot by the camera to judge whether the single fruit belongs to inedible categories; and performing classification feature extraction on the single fruits of the edible category, and grading the single fruits according to multi-feature indexes generated by feature extraction results.
Further, before determining whether the single fruit belongs to the inedible category, the method further comprises: and calibrating the shooting position of the camera by using the standard reference object, wherein the calibrated shooting position is set as the shooting distance with the minimum influence on the actual size of the single fruit.
Further, the classifying and characteristic extracting of the single fruits of the edible category comprises extracting size characteristics, fruit shape characteristics, color characteristics and defect characteristics of the single fruits.
Further, the extracting disease characteristics of the single-fruit image shot by the camera includes: reading an input single-effect image, converting the single-effect image from an RGB color space to an LAB color space by conversion processing, classifying color features by using Kmeans clustering, marking each pixel in the image as a result of Kmeans, generating an image segmented according to colors, and obtaining a color space range containing diseases.
Further, the determining whether the single fruit belongs to the inedible category includes: carrying out disease region identification on the color space range containing the diseases obtained after segmentation by utilizing a gray level co-occurrence matrix in combination with a Gabor filter, and reducing the dimension of the obtained characteristic parameters by adopting a principal component analysis method; diseases are distinguished through characteristic parameters of different diseases, and the characteristic parameters are input into an SVM classifier to generate a classification model.
Further, the diameter of the longitudinal section of the single fruit is calculated by using a minimum circumcircle method to serve as the extracted big and small characteristics; calculating the ratio of the sample cross section area of the single fruit to the minimum circumscribed circle area by using a minimum circumscribed circle method as the extracted fruit shape characteristic; using the color degree of the red component extracted under the HSV space as the extracted color characteristic; the area of the defect region obtained by the morphological operation of the image processing is used as the extracted defect feature.
Further, the single-effect image shot by the camera comprises a single-effect image shot directly by the camera and each angle image of the single-effect shot by the plane mirror.
Compared with the prior art, the invention has the following advantages:
the invention replaces manpower with computer vision, replaces a large amount of labor force and avoids subjectivity and fatigue of the manpower; the invention eliminates the inedible apples before grading work, and strives to realize a relatively complete system; the invention adopts multi-characteristic indexes for detection and classification, and improves the quality and efficiency compared with the traditional single characteristic; the invention simulates the operation on the actual production line, obtains the whole surface of the apple and improves the accuracy; the invention is classified according to the standard of 'quality grade requirement of fresh apples', and has more scientificity and foundation.
Based on the reasons, the invention can be widely popularized in the field of fruit classification.
Drawings
In order to more clearly illustrate the embodiments of the present invention 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 introduced below, and it is obvious that the drawings in the following description are 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 flow chart of the grading method of the present invention.
FIG. 2 is a schematic diagram of a grading apparatus according to an embodiment of the present invention.
Detailed Description
In order to make the technical solutions of the present invention better understood, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. 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.
As shown in fig. 1, the present invention provides an image-based apple objective grading method, including: extracting disease characteristics of the single fruit image shot by the camera to judge whether the single fruit belongs to inedible categories; and performing classification feature extraction on the single fruits of the edible category, and grading the single fruits according to multi-feature indexes generated by feature extraction results.
Further, before determining whether the single fruit belongs to the inedible category, the method further comprises: and calibrating the shooting position of the camera by using the standard reference object, wherein the calibrated shooting position is set as the shooting distance with the minimum influence on the actual size of the single fruit. Specifically, as an implementation of the calibration, a USB camera is used to take a picture of a single coin, and image processing is performed on the picture: firstly, graying processing is carried out, after binarization processing, the number of pixels occupied by the diameter of the coin is calculated, the actual diameter of the unitary coin is 25mm, and then the ratio of the actual diameter to the number of pixels occupied by the diameter can be obtained:
wherein M is the actual diameter length of the coin; n is the number of pixels occupied by the coin diameter. Through a plurality of tests, the measurement of the diameter of the same apple at different heights is calculated, and the fact that the height of the camera is 32cm, and the influence of the height on the actual size of an object is minimum is obtained.
According to the apple and pear standards, once the surface of an apple has rottenness and insect damage, the apple is not brought into the grading standards any more, so that a large number of rottenness apples and insect damage apple data sets are manufactured, and a Kmeans algorithm is used for dividing the rottenness apples and the insect damage apples to extract disease area parts. The method specifically comprises the following steps:
a) reading an input single-effect image, converting the image from RGB to LAB color space, classifying colors in 'a x b' space by using Kmeans clustering, marking each pixel in the image as a result of Kmeans, generating an image of dividing the image according to colors, and selecting a color space range containing diseases.
The Kmeans algorithm uses the Euclidean distance formula:
the distance between each pair of data objects can be calculated through a formula, and clustering is carried out according to the distance to form the designated class number K. Randomly taking k data as class centers, and carrying out a class center iteration process:
wherein, the CenterkIs heart-like; ckIs of the kth class.
There are generally two conditions for iteration termination: 1) reaching the specified iteration times; 2) the centroid is no longer changed.
b) And identifying the disease region by combining the segmented disease region with a gray level co-occurrence matrix and a Gabor filter, and reducing the dimension of the obtained characteristic parameters by adopting a principal component analysis method to remove redundant information. Diseases are distinguished through characteristic parameters of different diseases, the characteristic parameters are input into an SVM classifier, and a training model is generated through multiple tests.
The texture features are analyzed by utilizing a gray level co-occurrence matrix and a Gabor filter, wherein the gray level co-occurrence matrix reflects the comprehensive information of the image texture in the direction, the adjacent interval, the variation amplitude and the speed; the gray level co-occurrence matrix is obtained by counting the condition that two pixel points of the image which keep a certain fixed distance in a certain direction respectively have a certain gray level.
The elements in the gray level co-occurrence matrix are represented by probability values, i.e. dividing each element value by the sum of each element to obtain a normalized value of each element less than 1
Namely:
wherein i and j respectively represent the gray scales of two pixels; delta is the spatial position relation between two pixels; s is the sum of each element in the gray level co-occurrence matrix.
The surface image texture characteristic value table is three measures obtained by the apple surface gray level image based on a gray level co-occurrence matrix, namely energy, entropy and contrast:
wherein, ASM, ENT and CON are respectively the energy, entropy and contrast of surface texture features; l represents the gray level of the image.
In order to acquire complete surface information, the invention additionally considers a direction parameter theta representing the spatial position relationship and a distance parameter d representing the spatial position relationship, calculates a gray level co-occurrence matrix under the condition that theta is 0 degrees, 45 degrees, 90 degrees and 135 degrees and d is 1, and obtains the mean value and the variance of the obtained characteristic indexes as characteristic vectors.
At this time, the function of the two-dimensional Gabor transform is expressed as:
the parameters affecting the Gabor filter include the center frequency f, the direction angle θ, and the bandwidth of the filter determined by σ. The Gabor filter in the invention adopts 40 Gabor filters consisting of 5 central frequencies and 8 directions to process the image, and the sampling interval is set to be pi/8. The two-dimensional signals are filtered by the Gabor filters in different directions and scales, and the characteristics of the two-dimensional signals in a frequency domain space can be comprehensively reflected.
The principal component analysis method described above uses a new set of low-dimensional feature vectors to represent the original sample as accurately as possible, i.e., m-dimensional feature vectors Q ═ Q1, Q2,. and Q ] T are used to replace the original n-dimensional feature vectors P ═ P1, P2,. and Pn ] T (m < n).
The method adopts the principal component analysis method to remove the correlation among samples and realize the compression of data, and the dimension reduction method is adopted in the invention to reduce the dimension of the Gabor characteristic vector of the image from 40 dimensions to 10 dimensions.
The classification method specifically comprises the following steps: libsvm was used to classify into 3 classes, putrefaction, insect damage and others. The input vectors respectively correspond to the surface image texture feature vectors, and the training accuracy and the class labels are output; obtaining a proper classification mode by using enough samples through repeated training; inputting the texture feature vector of the surface image of the apple to be detected into the image to be detected, and obtaining a final classification result.
Further, the classifying and characteristic extracting of the single fruits of the edible category comprises extracting size characteristics, fruit shape characteristics, color characteristics and defect characteristics of the single fruits. Specifically, the method comprises the following steps:
when a certain apple sample is actually detected, as a preferred embodiment of the invention, the device shown in fig. 2 is utilized in a closed space, and two LED illuminating lamps are arranged on two sides of a camera; four plane mirrors are symmetrically and obliquely arranged to reflect the whole surface of the apple; several pictures of the apple can be obtained, the pictures are divided by a physical method, the whole surface of the apple can be detected, the pictures are divided by Kmeans, and the pictures are input into a trained model, so that whether the apple is rotten, damaged by insects or other apples is concluded. If the output is the apple with diseases, the apple is rejected.
Further, extracting size characteristics, fruit shape characteristics, color characteristics and defect characteristics of the disease-free apples, and dividing the obtained characteristic standard into special-grade fruits, first-grade fruits, second-grade fruits and other outer fruits according to the quality grade requirement of the fresh apples.
The extraction of the size characteristics is to calculate the diameter of the longitudinal section of the apple by using a minimum circumcircle method, and the fruit shape index is expressed by using the minimum circumcircle method as follows:
wherein S0 is the area of the cross-section of the apple sample; s1 is the area of the minimum circumscribed circle.
The color characteristic is that the color of the HSV space is less influenced by illumination, so the color related characteristic is researched in the HSV space, the H and S spaces are selected to extract the red component, and the color degree is expressed as follows:
wherein S is the number of pixels on the surface of the apple; s2 is the number of pixels of the red component.
The defect feature utilizes morphological operation of image processing to extract a defect area of a picture, firstly graying, secondly binarizing, then extracting an edge, filling the edge to obtain another edge, finally obtaining the difference between the two edges, namely the defect edge, and obtaining the pixel number of the defect after filling the defect edge. The area of the defect is expressed as:
Sd＝N*k
wherein N is the number of defective pixels; k is the ratio of the actual size to the pixel size.
The image-based apple objective grading method disclosed by the invention replaces manpower with computer vision, replaces a large amount of labor force and avoids subjectivity and fatigue of the manpower; before grading work, inedible apples are eliminated, and a relatively complete system is realized; the detection grading is carried out by adopting the multi-feature indexes, so that the quality and the efficiency are improved compared with the traditional single feature; the operation on the actual production line is simulated, the whole surface of the apple is obtained, and the accuracy is improved; grading according to the standard of 'quality grade requirement of fresh apples' has more scientificity and foundation.
Finally, it should be noted that: the above embodiments are only used to illustrate the technical solution of the present invention, and not to limit the same; while the invention has been described in detail and with reference to the foregoing embodiments, it will be understood by those skilled in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some or all of the technical features may be equivalently replaced; and the modifications or the substitutions do not make the essence of the corresponding technical solutions depart from the scope of the technical solutions of the embodiments of the present invention.
Claims (7)
1. An image-based apple objective grading method is characterized by comprising the following steps:
extracting disease characteristics of the single fruit image shot by the camera to judge whether the single fruit belongs to inedible categories;
and performing classification feature extraction on the single fruits of the edible category, and grading the single fruits according to multi-feature indexes generated by feature extraction results.
2. The image-based apple objective rating method of claim 1, wherein prior to determining whether the single fruit belongs to the inedible category further comprises:
and calibrating the shooting position of the camera by using the standard reference object, wherein the calibrated shooting position is set as the shooting distance with the minimum influence on the actual size of the single fruit.
3. The image-based apple objective rating method of claim 1 or 2, wherein the classifying feature extraction of the single fruits of the edible category comprises extracting size features, fruit shape features, color features and defect features of the single fruits.
4. The image-based apple objective grading method according to claim 1, wherein the disease feature extraction of the single-fruit image shot by the camera comprises:
reading an input single-effect image, converting the single-effect image from an RGB color space to an LAB color space by conversion processing, classifying color features by using Kmeans clustering, marking each pixel in the image as a result of Kmeans, generating an image segmented according to colors, and obtaining a color space range containing diseases.
5. The image-based apple objective rating method of claim 4, wherein said determining whether the single fruit belongs to the inedible category comprises:
carrying out disease region identification on the color space range containing the diseases obtained after segmentation by utilizing a gray level co-occurrence matrix in combination with a Gabor filter, and reducing the dimension of the obtained characteristic parameters by adopting a principal component analysis method;
diseases are distinguished through characteristic parameters of different diseases, and the characteristic parameters are input into an SVM classifier to generate a classification model.
6. The image-based apple objective rating method of claim 3,
calculating the diameter of the longitudinal section of the single fruit by using a minimum circumcircle method as the extracted big and small characteristics;
calculating the ratio of the sample cross section area of the single fruit to the minimum circumscribed circle area by using a minimum circumscribed circle method as the extracted fruit shape characteristic;
using the color degree of the red component extracted under the HSV space as the extracted color characteristic;
the area of the defect region obtained by the morphological operation of the image processing is used as the extracted defect feature.
7. The image-based apple objective grading method according to claim 1, wherein the single-fruit images shot by the camera comprise single-fruit images shot directly by the camera and single-fruit angle images shot by the camera and reflected by a plane mirror.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911158969.0A CN110929787B (en) | 2019-11-22 | 2019-11-22 | Apple objective grading system based on image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911158969.0A CN110929787B (en) | 2019-11-22 | 2019-11-22 | Apple objective grading system based on image |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110929787A true CN110929787A (en) | 2020-03-27 |
| CN110929787B CN110929787B (en) | 2024-02-06 |
Family
ID=69850836
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911158969.0A Active CN110929787B (en) | 2019-11-22 | 2019-11-22 | Apple objective grading system based on image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110929787B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112556606A (en) * | 2020-12-24 | 2021-03-26 | 宁夏农林科学院农业经济与信息技术研究所（宁夏农业科技图书馆） | Self-propelled wolfberry fruit actual measurement method and device based on binocular vision |
| CN113887389A (en) * | 2021-09-29 | 2022-01-04 | 湖南省博世康中医药有限公司 | Quality inspection and classification method for traditional Chinese medicine decoction pieces based on image recognition |
| TWI830982B (en) * | 2021-01-23 | 2024-02-01 | 鴻海精密工業股份有限公司 | Image recognition method based on texture classification, electronic device, and storage medium |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2012039597A2 (en) * | 2010-09-23 | 2012-03-29 | Universiti Putra Malaysia | Fruit ripeness grading system |
| CN110148122A (en) * | 2019-05-17 | 2019-08-20 | 南京东奇智能制造研究院有限公司 | Apple presentation quality stage division based on deep learning |
| CN110276386A (en) * | 2019-06-11 | 2019-09-24 | 济南大学 | A kind of apple grading method and system based on machine vision |
-
2019
- 2019-11-22 CN CN201911158969.0A patent/CN110929787B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2012039597A2 (en) * | 2010-09-23 | 2012-03-29 | Universiti Putra Malaysia | Fruit ripeness grading system |
| CN110148122A (en) * | 2019-05-17 | 2019-08-20 | 南京东奇智能制造研究院有限公司 | Apple presentation quality stage division based on deep learning |
| CN110276386A (en) * | 2019-06-11 | 2019-09-24 | 济南大学 | A kind of apple grading method and system based on machine vision |
Non-Patent Citations (1)
| Title |
|---|
| 李聪 等: ""基于计算机视觉的苹果自动分级方法研究"" * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112556606A (en) * | 2020-12-24 | 2021-03-26 | 宁夏农林科学院农业经济与信息技术研究所（宁夏农业科技图书馆） | Self-propelled wolfberry fruit actual measurement method and device based on binocular vision |
| TWI830982B (en) * | 2021-01-23 | 2024-02-01 | 鴻海精密工業股份有限公司 | Image recognition method based on texture classification, electronic device, and storage medium |
| CN113887389A (en) * | 2021-09-29 | 2022-01-04 | 湖南省博世康中医药有限公司 | Quality inspection and classification method for traditional Chinese medicine decoction pieces based on image recognition |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110929787B (en) | 2024-02-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Sahu et al. | Defect identification and maturity detection of mango fruits using image analysis | |
| CN109978822B (en) | Banana maturity judging modeling method and judging method based on machine vision | |
| CN109002851B (en) | Fruit classification method based on image multi-feature fusion and application | |
| CN110929787B (en) | Apple objective grading system based on image | |
| CN104990892B (en) | The spectrum picture Undamaged determination method for establishing model and seeds idenmtification method of seed | |
| Rahamathunnisa et al. | Vegetable disease detection using k-means clustering and svm | |
| Sahu et al. | Identification and classification of mango fruits using image processing | |
| CN106610969A (en) | Multimodal information-based video content auditing system and method | |
| Abdullah et al. | Automated inspection system for colour and shape grading of starfruit (Averrhoa carambola L.) using machine vision sensor | |
| CN107505325B (en) | Omnibearing quality detection method for winter jujube fruits | |
| CN109377487B (en) | Fruit surface defect detection method based on deep learning segmentation | |
| Ünal et al. | Classification of hazelnut kernels with deep learning | |
| CN113222959B (en) | Fresh jujube wormhole detection method based on hyperspectral image convolutional neural network | |
| CN105067532A (en) | Method for identifying early-stage disease spots of sclerotinia sclerotiorum and botrytis of rape | |
| CN103528967A (en) | Hyperspectral image based overripe Lonicera edulis fruit identification method | |
| CN109883967A (en) | A kind of Eriocheir sinensis quality grade method of discrimination based on information fusion | |
| Zhang et al. | Grasping point detection of randomly placed fruit cluster using adaptive morphology segmentation and principal component classification of multiple features | |
| CN113920376A (en) | Method for identifying wheat seed varieties based on light-weight convolutional neural network | |
| Yro et al. | Cocoa beans fermentation degree assessment for quality control using machine vision and multiclass svm classifier | |
| CN114863106B (en) | Centrifugal fertilizer spreading particle distribution detection method | |
| Jana et al. | Intra-class recognition of fruits using color and texture features with neural classifiers | |
| CN108073940A (en) | A kind of method of 3D object instance object detections in unstructured moving grids | |
| Ji et al. | Apple color automatic grading method based on machine vision | |
| Li et al. | Automatic recognition of tea bud image based on support vector machine | |
| CN116229182A (en) | Small particle coffee defect green bean detection method based on feature fusion |
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 |