Muutke küpsiste eelistusi

E-raamat: Computer Vision-Based Agriculture Engineering [Taylor & Francis e-raamat]

(College of Science and Information,Qingdao Agricultural University, Shandong, China)
  • Formaat: 330 pages, 75 Tables, black and white; 162 Illustrations, black and white
  • Ilmumisaeg: 30-Sep-2019
  • Kirjastus: CRC Press
  • ISBN-13: 9780429289460
Teised raamatud teemal:
  • Taylor & Francis e-raamat
  • Hind: 170,80 €*
  • * hind, mis tagab piiramatu üheaegsete kasutajate arvuga ligipääsu piiramatuks ajaks
  • Tavahind: 244,00 €
  • Säästad 30%
Computer Vision-Based Agriculture EngineeringSuurem pilt
  • Formaat: 330 pages, 75 Tables, black and white; 162 Illustrations, black and white
  • Ilmumisaeg: 30-Sep-2019
  • Kirjastus: CRC Press
  • ISBN-13: 9780429289460
Teised raamatud teemal:
Taylor & Francis e-raamatudRohkem infot Taylor & Francis e-raamatute kohta
Raamatu kodulehekülg: https://www.taylorfrancis.com/books/9780429289460
In recent years, computer vision is a fast-growing technique of agricultural engineering, especially in quality detection of agricultural products and food safety testing. It can provide objective, rapid, non-contact and non-destructive methods by extracting quantitative information from digital images. Significant scientific and technological advances have been made in quality inspection, classification and evaluation of a wide range of food and agricultural products. Computer Vision-Based Agriculture Engineering focuses on these advances.The book contains 25 chapters divided into six parts. It covers computer vision, image processing, hyperspectral imaging and other related technologies in peanut aflatoxin, peanut and corn quality varieties, carrot and potato quality, as well as pest and disease detection.Features:Discusses various detection methods in a variety of agricultural crops.Each chapter includes material and methods used, results and analysis, and discussion with conclusions.Covers basic theory, technical methods, and engineering cases.Provides comprehensive coverage on methods of variety identification, quality detection, and detection of key indicators of agricultural products safetyPresents information on technology of artificial intelligence including deep learning and transfer learning.Computer Vision-Based Agriculture Engineering is a summary of the authors work over the past 10 years. Professor Han has presented his most recent research results in the book of all 25 chapters. This unique work provides student, engineers and technologists working in research, development, and operations in the agricultural engineering with critical, comprehensive and readily accessible information. The book applies development of artificial intelligence theory and methods including depth learning and transfer learning to the field of agricultural engineering testing.
Preface xv
Author xvii
Chapter 1 Detecting Anatoxin in Agricultural Products by Hyperspectral Imaging: A Review
1(28)
1.1 Introduction
1(2)
1.2 Main Detecting Methods
3(2)
1.2.1 Hyperspectral Imaging (HSI)
3(1)
1.2.2 Near-Infrared Spectroscopy (NIRS)
3(2)
1.2.3 Flow Chart Showing Typical Steps
5(1)
1.3 Detection of Anatoxin in Agricultural Products
5(13)
1.3.1 Corn or Maize
5(1)
1.3.1.1 Detection of Anatoxin by Hyperspectral Images
5(4)
1.3.1.2 Detection of Anatoxin by Other Methods
9(3)
1.3.1.3 Detect Other Fungi
12(1)
1.3.2 Cereals, Nuts, and Others
13(1)
1.3.2.1 Wheat, Barley, and Rice
13(1)
1.3.2.2 Pistachio Nuts, Hazelnuts, Brazil Nuts, and Peanuts
14(2)
1.3.2.3 Chili Pepper
16(2)
1.4 Limitation and Future Trends
18(2)
1.4.1 Limitation
18(1)
1.4.2 Future Trends
19(1)
1.5 Conclusions
20(9)
References
20(9)
Chapter 2 Anatoxin Detection by Fluorescence Index and Narrowband Spectra Based on Hyperspectral Imaging
29(12)
2.1 Introduction
29(1)
2.2 Experiment Materials
30(3)
2.2.1 Sample Preparation and Image Acquisition
30(2)
2.2.2 Illumination Compensation and Kernel Segmentation
32(1)
2.3 Data Processing and Result Analysis
33(4)
2.3.1 Fluorescence Index
33(1)
2.3.2 Recognition and Regression
34(1)
2.3.3 Narrowband Spectra
35(2)
2.4 Discussion
37(2)
2.5 Conclusions
39(2)
References
39(2)
Chapter 3 Application-Driven Key Wavelength Mining Method for Anatoxin Detection Using Hyperspectral Data
41(16)
3.1 Introduction
41(1)
3.2 Materials
42(2)
3.2.1 Experiment Materials
42(1)
3.2.2 System Integration
43(1)
3.3 Methods
44(3)
3.3.1 Data Preprocessing
44(2)
3.3.2 Recognition Methods
46(1)
3.4 Results
47(4)
3.4.1 Hyperspectral Wave by ASD
47(1)
3.4.2 Multispectral Images by Liquid Crystal Tunable Filter (HTLF)
48(1)
3.4.3 Hyperspectral Images by GSM
49(2)
3.5 Discussion
51(3)
3.5.1 Key Wavelengths Selected by Weighted Voting
51(2)
3.5.2 Sorter Design
53(1)
3.6 Conclusion
54(3)
References
55(2)
Chapter 4 Deep Learning-Based Anatoxin Detection of Hyperspectral Data
57(12)
4.1 Introduction
57(2)
4.2 Materials and Methods
59(3)
4.2.1 Peanut Sample Preparation
59(1)
4.2.2 Hyperspectral Imaging System and Image Acquisition
59(1)
4.2.3 Hyperspectral Imaging Preprocessing
60(1)
4.2.4 CNN of Deep Learning Method
61(1)
4.3 Results and Discussion
62(3)
4.3.1 Anatoxin Detection Using Key Band Images
62(1)
4.3.2 Anatoxin Detection Using Spectral and Images
63(2)
4.4 Conclusion
65(4)
References
66(3)
Chapter 5 Pixel-Level Anatoxin Detection Based on Deep Learning and Hyperspectral Imaging
69(14)
5.1 Introduction
69(2)
5.2 Materials and Methods
71(3)
5.2.1 Peanut Sample Preparation
71(1)
5.2.2 Hyperspectral Imaging System and Image Acquisition
71(1)
5.2.3 Hyperspectral Imaging Preprocessing
72(1)
5.2.4 CNN of Deep Learning Method
73(1)
5.3 Results and Discussion
74(5)
5.3.1 Deep Learning for Training Kernels
74(1)
5.3.2 Deep Learning for Testing Kernels
75(2)
5.3.3 Models Compared for All Kernels
77(2)
5.4 Discussion
79(1)
5.5 Conclusions
80(3)
References
80(3)
Chapter 6 A Method of Detecting Peanut Cultivars and Quality Based on the Appearance Characteristic Recognition
83(8)
6.1 Introduction
83(1)
6.2 Materials and Method
84(3)
6.2.1 Materials for Test
84(1)
6.2.2 Image Acquisition and Pretreatment
84(2)
6.2.3 The Appearance Characteristic Index of the Seed
86(1)
6.2.4 The Establishment of the Recognition Model
86(1)
6.3 Results and Analysis
87(1)
6.3.1 The Result of Recognition on Peanut Varieties
87(1)
6.3.2 The Result of Recognition on Peanut Qualities
88(1)
6.4 Analysis of the Results of Recognition and Detection
88(1)
6.5 Discussions
89(2)
References
89(2)
Chapter 7 Quality Grade Testing of Peanut Based on Image Processing
91(8)
7.1 Introduction
91(1)
7.2 Materials and Method
92(3)
7.2.1 Materials for Testing
92(1)
7.2.2 Maintaining the Integrity of the Specifications
93(1)
7.2.3 Model of Quality Recognition
94(1)
7.2.4 Method of Grading
94(1)
7.3 Results and Analysis
95(2)
7.3.1 Analysis of Recognition Results of Grains' Quality
95(1)
7.3.2 Analysis of the Result of Specification and Grading
96(1)
7.4 Conclusions
97(2)
References
97(2)
Chapter 8 Study on Origin Traceability of Peanut Pods Based on Image Recognition
99(8)
8.1 Introduction
99(1)
8.2 Materials and Method
100(2)
8.2.1 Test Materials
100(1)
8.2.2 Methods
101(1)
8.2.3 The Method Used to Optimize
101(1)
8.3 Results and Analysis
102(1)
8.4 Discussions
103(1)
8.5 Conclusion
104(3)
References
104(3)
Chapter 9 Study on the Pedigree Clustering of Peanut Pod's Variety Based on Image Processing
107(10)
9.1 Introduction
107(1)
9.2 Materials and Method
108(2)
9.2.1 Experimental Materials
108(1)
9.2.2 Methods
109(1)
9.2.2.1 PCA Algorithm
110(1)
9.2.2.2 Clustering Algorithm
110(1)
9.3 Conclusion and Analysis
110(3)
9.3.1 Statistical Characteristics Clustering
110(1)
9.3.2 PCA Clustering
111(2)
9.4 Discussions
113(1)
9.5 Conclusions
114(3)
References
114(3)
Chapter 10 Image Features and DUS Testing Traits for Identification and Pedigree Analysis of Peanut Pod Varieties
117(14)
10.1 Introduction
117(1)
10.2 Materials and Method
118(4)
10.1.1 Materials
118(1)
10.1.1.1 Peanut Samples
118(1)
10.1.1.2 Image Acquisition
119(1)
10.1.2 Methods
119(1)
10.1.2.1 Feature Extraction
119(2)
10.1.2.2 Analysis and Identification Model
121(1)
10.3 Results and Analysis
122(4)
10.3.1 Feature Selection by Fisher
122(1)
10.3.2 Variety Identification by SVM
123(2)
10.3.3 Paternity Analysis by E-Means
125(1)
10.4 Discussions
126(2)
10.4.1 Biological Basis for Seed Testing with Appearance
126(1)
10.4.2 Finding Candidate Features for DUS Testing
127(1)
10.5 Conclusions
128(3)
References
128(3)
Chapter 11 Counting Ear Rows in Maize Using Image Processing Method
131(6)
11.1 Introduction
131(1)
11.2 Materials
132(1)
11.3 Procedure
132(2)
11.3.1 Image Obtaining
132(1)
11.3.2 Characteristic Indicators of Corn Varieties
132(1)
11.3.3 Pretreatment
132(1)
11.3.4 Construction of the Counting Model
132(2)
11.4 Results and Analysis
134(1)
11.5 Conclusions
135(2)
References
135(2)
Chapter 12 Single-Seed Precise Sowing of Maize Using Computer Simulation
137(14)
12.1 Introduction
137(2)
12.2 Materials and Methods
139(4)
12.2.1 Mathematical Depiction of the Problem
139(2)
12.2.2 Method of Analog Simulation
141(1)
12.2.2.1 Computer Simulation of Planting Method
141(1)
12.2.2.2 Seedling Missing Spots and Missing Seedling Compensation
142(1)
12.3 Results and Analysis
143(5)
12.3.1 Comparison of the Two Planting Methods' Yield
143(2)
12.3.2 Influence of Field Seedling Emergence Rate on Yield
145(1)
12.3.3 Interactions between Sensitivity and Field Seedling Emergence Rate
146(1)
12.3.4 Seedling Missing Spots and Its Distribution Rule
146(2)
12.4 Conclusion
148(3)
References
149(2)
Chapter 13 Identifying Maize Surface and Species by Transfer Learning
151(16)
13.1 Introduction
151(2)
13.2 Materials and Method
153(2)
13.2.1 Image Characteristics
153(1)
13.2.2 Workflow Diagram
153(1)
13.2.3 Convolutional Neural Network
153(2)
13.2.4 Transfer Learning
155(1)
13.3 Performance of the Model
155(2)
13.4 Performance of Limited Data Model
157(1)
13.5 Comparison with Manual Method
157(1)
13.6 Expanding Application
158(2)
13.7 Discussion
160(3)
13.8 Conclusions and Future Work
163(4)
References
163(4)
Chapter 14 A Carrot Sorting System Using Machine Vision Technique
167(16)
14.1 Introduction
167(1)
14.2 Materials and Methods
168(7)
14.2.1 Carrot Samples
168(1)
14.2.2 Grading System
169(1)
14.2.3 Image Processing and Detection Algorithms
170(1)
14.2.3.1 Image Preprocessing and Segmentation
170(1)
14.2.3.2 Shape Detection Algorithm
171(1)
14.2.3.3 Fibrous Root Detection
172(3)
14.2.3.4 Surface Crack Detection
175(1)
14.2.4 Bayes Classifier
175(1)
14.3 Experimental Results
175(3)
14.3.1 Detection
175(1)
14.3.2 Fibrous Root Detection
176(1)
14.3.3 Crack Detection
177(1)
14.3.4 Time Efficiency
178(1)
14.4 Discussions
178(2)
14.4.1 Shape Detection
178(1)
14.4.2 Fibrous Root Detection
178(1)
14.4.3 Crack Detection
179(1)
14.4.4 Future Work
179(1)
14.5 Conclusions
180(3)
References
180(3)
Chapter 15 A New Automatic Carrot Grading System Based on Computer Vision
183(14)
15.1 Introduction
183(1)
15.2 Materials and Methods
184(7)
15.2.1 Materials
184(1)
15.2.2 Design of Carrot Grading System
184(2)
15.2.3 Image Processing and Grading Algorithms
186(1)
15.2.3.1 Image Acquisition
186(1)
15.2.3.2 Image Preprocessing
187(1)
15.2.3.3 Defect Detection Algorithms
188(2)
15.2.3.4 Grading Regular Carrots by Size
190(1)
15.2.4 Control of Grading System
190(1)
15.3 Results
191(2)
15.3.1 Defect Detection
191(1)
15.3.2 Regular Carrot Grading
191(1)
15.3.3 Time Efficiency
191(1)
15.3.4 Performance Parameter
192(1)
15.4 Discussion
193(1)
15.5 Conclusion
194(3)
References
194(3)
Chapter 16 Identifying Carrot Appearance Quality by Transfer Learning
197(16)
16.1 Introduction
197(2)
16.2 Materials and Method
199(2)
16.2.1 Image Characteristics
199(1)
16.2.2 Workflow Diagram
199(1)
16.2.3 Convolutional Neural Network
199(2)
16.2.4 Transfer Learning
201(1)
16.3 Performance of the Model
201(2)
16.4 Comparison with Manual Work
203(3)
16.5 An Expanding Application
206(1)
16.6 Discussion
207(2)
16.7 Conclusions and Future Work
209(4)
References
210(3)
Chapter 17 Grading System of Pear's Appearance Quality Based on Computer Vision
213(12)
17.1 Introduction
213(1)
17.2 System Development
214(3)
17.2.1 Formation of Grading System's Hardware Device
214(1)
17.2.2 Concrete Implement
215(1)
17.2.3 Testing of Real Object
215(1)
17.2.4 Software System
216(1)
17.3 Implementation of Algorithm
217(4)
17.3.1 Detection of Defects
217(2)
17.3.2 Feature Extraction
219(1)
17.3.3 Grade Judgment
219(2)
17.4 Results and Discussion
221(2)
17.5 Ending Words
223(2)
References
223(2)
Chapter 18 Study on Defect Extraction of Pears with Rich Spots and Neural Network Grading Method
225(10)
18.1 Introduction
225(1)
18.2 Material and Method
226(4)
18.2.1 Experimental Materials
226(1)
18.2.2 Image Preprocessing
227(1)
18.2.2.1 Background Removing and Outline Extraction
227(1)
18.2.2.2 Removal of Spots on the Surface
228(1)
18.2.2.3 Extraction of Defective Parts
229(1)
18.2.3 Feature Extraction and Recognition
229(1)
18.2.3.1 Scalarization of National Standard
229(1)
18.2.3.2 Extraction of Fruit Type and Defective Part
230(1)
18.2.3.3 Judgment of ANN Grade
230(1)
18.3 Results and Analysis
230(3)
18.3.1 Effect of Spot Removal and Defect Extraction
230(1)
18.3.2 Grade Judgment of Fruit Type and Defect
231(1)
18.3.3 Comprehensive Grade Judgments
231(1)
18.3.4 Influencing Factors of Grading of Comprehensive Quality
231(2)
18.4 Discussions
233(2)
References
233(2)
Chapter 19 Food Detection Using Infrared Spectroscopy with k-ICA and k-SVM: Variety, Brand, Origin, and Adulteration
235(12)
19.1 Introduction
235(2)
19.2 Materials and Methods
237(3)
19.2.1 Materials
237(1)
19.2.2 Algorithm
238(2)
19.2.3 Flowchart
240(1)
19.3 Results and Discussion
240(5)
19.3.1 Different Features Selection Method
241(1)
19.3.2 Different Recognition Models
242(1)
19.3.3 Different Samples or Features
243(2)
19.4 Summary
245(2)
References
245(2)
Chapter 20 Study on Vegetable Seed Electrophoresis Image Classification Method
247(10)
20.1 Introduction
247(1)
20.2 Material and Method
248(2)
20.2.1 Experimental Materials
248(1)
20.2.2 Method
248(2)
20.3 Results and Analysis
250(4)
20.3.1 Recognition of Corp
250(2)
20.3.2 Cluster Analysis
252(2)
20.4 Discussions
254(1)
20.5 Conclusions
255(2)
References
256(1)
Chapter 21 Identifying the Change Process of a Fresh Pepper by Transfer Learning
257(14)
21.1 Introduction
257(1)
21.2 Materials and Method
258(3)
21.2.1 Image Characteristics
258(1)
21.2.2 Workflow Diagram
259(1)
21.2.3 Convolutional Neural Network
259(1)
21.2.4 Transfer Learning
260(1)
21.3 The Performance of the Model
261(2)
21.4 Computer vs Humans
263(1)
21.5 Expanding Application
264(1)
21.6 Discussion
265(2)
21.7 Conclusions and Future Work
267(4)
References
267(4)
Chapter 22 Identifying the Change Process of Fresh Banana by Transfer Learning
271(12)
22.1 Introduction
271(1)
22.2 Materials and Method
272(4)
22.2.1 Image Dataset
272(2)
22.2.2 Convolutional Neural Network
274(1)
22.2.3 Transfer Learning
274(1)
22.2.4 Experimental Setup
275(1)
22.3 The Performance of Model
276(3)
22.3.1 Experimental Result
276(1)
22.3.2 Computer vs Humans
277(1)
22.3.3 Expanding Application
277(2)
22.4 Discussion
279(1)
22.5 Conclusions
280(3)
References
280(3)
Chapter 23 Pest Recognition Using Transfer Learning
283(12)
23.1 Introduction
283(1)
23.2 Materials and Methods
284(2)
23.2.1 Materials
284(1)
23.2.2 Introduce of the Model
285(1)
23.3 Result and Analysis
286(5)
23.3.1 Pests Recognition Result by Transfer Learning Model
286(1)
23.3.2 Comparison of the Model with Traditional Methods
286(1)
23.3.3 Comparison of the Model with Human Expert
286(2)
23.3.4 Universal of the Transfer Learning Model
288(3)
23.4 Discussion
291(2)
23.4.1 Image Numbers and Image Capture Environment
291(1)
23.4.2 Image Background and Segmentation
291(1)
23.4.3 Similar Outline Disturb
291(2)
23.5 Conclusion
293(2)
References
293(2)
Chapter 24 Using Deep Learning for Image-Based Plant Disease Detection
295(12)
24.1 Introduction
295(1)
24.2 Materials and Methods
296(3)
24.2.1 Dataset
296(1)
24.2.2 Introduced Model
297(2)
24.2.3 Experiment Environment
299(1)
24.2.4 Experiment Result
299(1)
24.3 Model Universal Adaptability
299(3)
24.3.1 Big Datasets Validation
299(1)
24.3.2 Small Datasets Validation
300(1)
24.3.3 Artificial Recognition
301(1)
24.4 Discussion
302(5)
24.4.1 Effect of Data Size
302(1)
24.4.2 Image Background
303(1)
24.4.3 Symptom Variations
303(1)
24.4.4 Machine Recognition and Human Recognition
304(1)
References
304(3)
Chapter 25 Research on the Behavior Trajectory of Ornamental Fish Based on Computer Vision
307(14)
25.1 Introduction
307(2)
25.2 Experimental Materials and Methods
309(4)
25.2.1 The Experimental Device
309(1)
25.2.2 Preprocessing
310(1)
25.2.3 The Positioning of the Fish
310(1)
25.2.4 Reduction of Actual 3D Coordinates
311(2)
25.3 Three-Dimensional Trajectory Analysis
313(5)
25.4 Discussion
318(1)
25.5 Conclusion
319(2)
References
319(2)
Index 321
Han Zhongzhi (1981.6-), Ph. D., Male, Born in Junan County, Shandong Province, China. Full professor of Qingdao Agricultural University, Master's supervisor, 3-level candidate of "1361" talent engineering, head of modern agricultural intelligent equipment innovation team; Chief expert of Qingdao agricultural intelligent equipment expert workstation, Evaluation expert of National Natural Science Foundation and National IOT Special Fund, Intel Certified Visual Computing Engineer, member of International Computer Association (ACM) Expert Committee, Reviewer of many journals such as "Computers and Electronics in Agriculture" and Editorial Committee of "Higher Education Research and Practice". The main research area is computer vision intelligent detection in agricultural products.
Püsilink: https://www.kriso.ee/db/9780429289460_pe.html
Märksõnad: