Application of Deep and Machine Learning Using Image Analysis to Detect Fungal Contamination of Rapeseed
Abstract
:1. Introduction
2. Materials and Methods
2.1. Preparation of Samples
2.2. Image Preparation by Microscope
2.3. Image Processing by Software
2.4. The Structure of the Neural Network
- Input layer, i.e., 685 × 685 × 4 bitmap, in which 32-bit primary images in RGBA format were included;
- Convolution layers for each loaded image, 8, 16, 32, or 64 filters were used. The kernel size in the first two convolution layers was 5 × 5, and 3 × 3 in the next two. Parameter “strides” determined the dimension of the convolution step (the distance of the convolution window that is moved) expressed as (width, height) was set as value (2, 2) [44,45];
- The pooling layer was used to reduce information included in previously acquired images. Operation MaxPooling allowed for the reduction in the resolution of bitmaps (learning case). For each map of features, 2 × 2-pixel areas were singled out. For each region, the maximum pool algorithm was used, i.e., the resultant pixel with the highest value was selected, and other pixels were omitted. Each operation resulted in image reduction in the outermost areas of this map. The “padding” parameter was set as “same”. It means the outermost areas of the matrix field that was processed were added after each MaxPooling operation in order to avoid excess data loss. As a result of this, the data size of the operation was the same as the size of input data. Convolution depth was doubled for each two-fold width and height reduction of the analyzed image. It minimized the problem of fitting networks to training sets via gradual resolution reduction of the processed image during tensor operations. Reduction of resolution on the stage of initial data processing could lead to the loss of some data and would affect the prediction abilities of the created network;
- A thick (fully connected) layer including 288 neurons with activation function “relu”, and one neuron with activation function “sigmoid”. The function of activating a rectified linear unit, causes it to return the standard activation ReLu with default values within the range: max (y, 0), for which basic maximum equals 0 and value “y” is treated as an input tensor. On the other hand, for low values (<=5) sigmoid returns a value close to zero, for high values (>5) the result of the function is close to one.
3. Results and Discussion
3.1. Results of Machine Learning
3.2. Results of Deep Learning
4. Conclusions
Author Contributions
Funding
Conflicts of Interest
References
- Gawrysiak-Witulska, M.; Siger, A.; Nogala-Kalucka, M. Degradation of tocopherols during near-ambient rapeseed drying. J. Food Lipids 2009, 16, 524–539. [Google Scholar] [CrossRef]
- Szydłowska-Czerniak, A.; Bartkowiak-Broda, I.; Karlović, I.; Karlovits, G.; Szłyk, E. Antioxidant capacity, total phenolics, glucosinolates and colour parameters of rapeseed cultivars. Food Chem. 2011, 127, 556–563. [Google Scholar] [CrossRef] [PubMed]
- Gawrysiak-Witulska, M.; Siger, A.; Rusinek, R. Degradation of tocopherols during rapeseed storage in simulated conditions of industrial silos. Int. Agrophys. 2016, 30, 39–45. [Google Scholar] [CrossRef]
- Wawrzyniak, J.; Gawrysiak-Witulska, M.; Rudzińska, M. Dynamics of phytosterol degradation in a bulk of rapeseed stored under different temperature and humidity conditions. J. Stored Prod. Res. 2019, 83, 292–304. [Google Scholar] [CrossRef]
- Hofius, D.; Sonnewald, U. Vitamin E biosynthesis: Biochemistry meets cell biology. Trends Plant Sci. 2003, 8, 6–8. [Google Scholar] [PubMed]
- Unal, H.; Sincik, M.; Izli, N. Comparison of some engineering properties of rapeseed cultivars. Ind. Crops Prod. 2009, 30, 131–136. [Google Scholar] [CrossRef]
- Bajpai, D.; Tyagi, V.K. Biodiesel: Source, Production, Composition, Properties and Its Benefits. J. Oleo Sci. 2006, 55, 487–502. [Google Scholar] [CrossRef] [Green Version]
- Kurtulmuş, F.; Ünal, H. Discriminating rapeseed varieties using computer vision and machine learning. Expert Syst. Appl. 2015, 42, 1880–1891. [Google Scholar] [CrossRef]
- Encinar, J.M.; Pardal, A.; Sánchez, N.; Nogales, S. Biodiesel by transesterification of rapeseed oil using ultrasound: A kinetic study of base-catalysed reactions. Energies 2018, 11, 229. [Google Scholar]
- Santaraite, M.; Sendzikiene, E.; Makareviciene, V.; Kazancev, K. Biodiesel production by lipase-catalyzed in situ transesterification of rapeseed oil containing a high free fatty acid content with ethanol in diesel fuel media. Energies 2020, 13, 2588. [Google Scholar] [CrossRef]
- Encinar, J.M.; Nogales-Delgado, S.; Sánchez, N.; González, J.F. Biolubricants from rapeseed and castor oil transesterification by using titanium isopropoxide as a catalyst: Production and characterization. Catalysts 2020, 10, 366. [Google Scholar] [CrossRef] [Green Version]
- Gawrysiak-Witulska, M.; Siger, A.; Wawrzyniak, J.; Nogala-Kalucka, M. Changes in Tocochromanol Content in Seeds of Brassica napus L. During Adverse Conditions of Storage. J. Am. Oil Chem. Soc. 2011, 88, 1379–1385. [Google Scholar] [CrossRef]
- Kasprzycka, A.; Skiba, K.; Tys, J. Influence of storage conditions on microbial quality of rapeseed cake and middlings. Int. Agrophys. 2010, 24, 261–265. [Google Scholar]
- Rusinek, R.; Kobyłka, R. Experimental study and discrete element method modeling of temperature distributions in rapeseed stored in a model bin. J. Stored Prod. Res. 2014, 59, 254–259. [Google Scholar] [CrossRef]
- Sankaran, S.; Mishra, A.; Ehsani, R.; Davis, C. A review of advanced techniques for detecting plant diseases. Comput. Electron. Agric. 2010, 72, 1–13. [Google Scholar] [CrossRef]
- Zhao, Y.-R.; Yu, K.-Q.; Li, X.; He, Y. Detection of Fungus Infection on Petals of Rapeseed (Brassica napus L.) Using NIR Hyperspectral Imaging. Sci. Rep. 2016, 6, 38878. [Google Scholar]
- Wawrzyniak, J.; Gawrysiak-Witulska, M.; Ryniecki, A. Management Control Points Related to the Lag Phase of Fungal Growth in a Stored Rapeseed Ecosystem. J. Am. Oil Chem. Soc. 2018, 95, 1223–1235. [Google Scholar] [CrossRef]
- Janda, K.; Markowska-Szczupak, A. Relationships Between Fungal Contamination and Some Physicochemical Properties of Rapeseeds. Ekologia 2015, 34, 65–71. [Google Scholar] [CrossRef] [Green Version]
- Przybył, K.; Gawałek, J.; Koszela, K.; Wawrzyniak, J.; Gierz, L. Artificial neural networks and electron microscopy to evaluate the quality of fruit and vegetable spray-dried powders. Case study: Strawberry powder. Comput. Electron. Agric. 2018, 155, 314–323. [Google Scholar]
- Koszela, K.; Otrząsek, J.; Zaborowicz, M.; Boniecki, P.; Mueller, W.; Raba, B.; Lewicki, A.; Przybył, K. Quality assessment of microwave-vacuum dried material with the use of computer image analysis and neural model. In Proceedings of the International Society for Optical Engineering, Athens, Greece, 5–6 April 2014. [Google Scholar]
- Ma, H.; Liu, Y.; Ren, Y.; Wang, D.; Yu, L.; Yu, J. Improved CNN Classification Method for Groups of Buildings Damaged by Earthquake, Based on High Resolution Remote Sensing Images. Remote Sens. 2020, 12, 260. [Google Scholar] [CrossRef] [Green Version]
- Przybył, K.; Duda, A.; Koszela, K.; Stangierski, J.; Polarczyk, M.; Gierz, Ł. Classification of Dried Strawberry by the Analysis of the Acoustic Sound with Artificial Neural Networks. Sensors 2020, 20, 499. [Google Scholar] [CrossRef] [PubMed] [Green Version]
- Przybył, K.; Boniecki, P.; Koszela, K.; Gierz, Ł.; Łukomski, M. Computer vision and artificial neural network techniques for classification of damage in potatoes during the storage process. Czech J. Food Sci. 2019, 37, 135–140. [Google Scholar]
- Boniecki, P.; Zaborowicz, M.; Pilarska, A.; Piekarska-Boniecka, H. Identification Process of Selected Graphic Features Apple Tree Pests by Neural Models Type MLP, RBF and DNN. Agriculture 2020, 10, 218. [Google Scholar] [CrossRef]
- Przybył, K.; Gawałek, J.; Koszela, K. Application of artificial neural network for the quality-based classification of spray-dried rhubarb juice powders. J. Food Sci. Technol. 2020. [Google Scholar] [CrossRef]
- Nowakowski, K.; Boniecki, P.; Tomczak, R.J.; Kujawa, S.; Raba, B. Identification of malting barley varieties using computer image analysis and artificial neural networks. In Proceedings of the SPIE—The International Society for Optical Engineering, Kuala Lumpur, Malaysia, 7–8 April 2012. [Google Scholar]
- Kujawa, S.; Mazurkiewicz, J.; Mueller, W.; Gierz, Ł.; Przybył, K.; Wojcieszak, D.; Zaborowicz, M.; Koszela, K.; Boniecki, P. Identification of co-substrate composted with sewage sludge using convolutional neural networks. In Proceedings of the Eleventh International Conference on Digital Image Processing, Guangzhou, China, 10–13 May 2019. [Google Scholar]
- Bauer, A.; Bostrom, A.G.; Ball, J.; Applegate, C.; Cheng, T.; Laycock, S.; Rojas, S.M.; Kirwan, J.; Zhou, J. Combining computer vision and deep learning to enable ultra-scale aerial phenotyping and precision agriculture: A case study of lettuce production. Hortic. Res. 2019, 6, 70. [Google Scholar] [CrossRef] [PubMed] [Green Version]
- Jiang, X.; Wang, Y.; Liu, W.; Li, S.; Liu, J. CapsNet, CNN, FCN: Comparative Performance Evaluation for Image Classification. Int. J. Mach. Learn. Comput. 2019, 9, 840–848. [Google Scholar] [CrossRef]
- Das, S. CNN Architectures: LeNet, AlexNet, VGG, GoogLeNet, ResNet and more. Medium 2017. Available online: https://medium.com/analytics-vidhya/cnns-architectures-lenet-alexnet-vgg-googlenet-resnet-and-more-666091488df5 (accessed on 4 September 2020).
- Liu, C.; Sun, W.; Chao, W.; Che, W. Convolution neural network for relation extraction. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Springer: Berlin/Heidelberg, Germany, 2013; Volume 8347, pp. 231–242. ISBN 9783642539169. [Google Scholar]
- Shawky, O.A.; Hagag, A.; El-Dahshan, E.-S.A.; Ismail, M.A. Remote sensing image scene classification using CNN-MLP with data augmentation. Optik 2020, 221, 165356. [Google Scholar] [CrossRef]
- Wawrzyniak, J.; Ryniecki, A.; Gawrysiak-Witulska, M. Kinetics of mould growth in the stored barley ecosystem contaminated with Aspergillus westerdijkiae, Penicillium viridicatum and Fusarium poae at 23–30 °C. J. Sci. Food Agric. 2013, 93, 895–901. [Google Scholar]
- ASABE. ASABE Standards D245.5. In Moisture Relationships of Plant-Based Agricultural Products; American Society of Agricultural Engineers: St. Joseph, MI, USA, 2007; pp. 538–550. [Google Scholar]
- Muthukrishnan, R.; Radha, M. Edge Detection Techniques for Image Segmentation. Int. J. Comput. Sci. Inf. Technol. 2011, 3, 259–267. [Google Scholar] [CrossRef]
- Chen, Y.; Li, X.; Cheng, Z.; Cheng, Y.; Zhai, X. Multidirectional edge detection based on gradient ghost imaging. Optik 2020, 207, 163768. [Google Scholar] [CrossRef]
- Przybył, K.; Ryniecki, A.; Niedbała, G.; Mueller, W.; Boniecki, P.; Zaborowicz, M.; Koszela, K.; Kujawa, S.; Kozłowski, R.J. Software supporting definition and extraction of the quality parameters of potatoes by using image analysis. In Proceedings of the Eighth International Conference on Digital Image Processing, Chengdu, China, 20–22 May 2016. [Google Scholar]
- Haralick, R.M.; Shanmugam, K.; Dinstein, I. Textural Features for Image Classification. IEEE Trans. Syst. Man. Cybern. 1973, SMC-3, 610–621. [Google Scholar] [CrossRef] [Green Version]
- Clausi, D.A. An analysis of co-occurrence texture statistics as a function of grey level quantization. Can. J. Remote Sens. 2002, 28, 45–62. [Google Scholar]
- Unser, M. Sum and Difference Histograms for Texture Classification. IEEE Trans. Pattern Anal. Mach. Intell. 1986, PAMI-8, 118–125. [Google Scholar] [CrossRef]
- Soh, L.-K.; Tsatsoulis, C. Texture analysis of SAR sea ice imagery using gray level co-occurrence matrices. IEEE Trans. Geosci. Remote Sens. 1999, 37, 780–795. [Google Scholar] [CrossRef] [Green Version]
- Lin, H.; Zhao, J.; Chen, Q.; Cai, J.; Zhou, P. Eggshell crack detection based on acoustic impulse response and supervised pattern recognition. Czech J. Food Sci. 2009, 27, 393–402. [Google Scholar] [CrossRef] [Green Version]
- Park, B.; Chen, Y.R. AE—Automation and Emerging Technologies: Co-occurrence Matrix Texture Features of Multi-spectral Images on Poultry Carcasses. J. Agric. Eng. Res. 2001, 78, 127–139. [Google Scholar]
- Briggs, F.H.; Bell, J.F.; Kesteven, M.J. Removing Radio Interference from Contaminated Astronomical Spectra Using an Independent Reference Signal and Closure Relations. Astron. J. 2000, 120, 3351–3361. [Google Scholar] [CrossRef] [Green Version]
- Nelli, F.; Nelli, F. Deep Learning with TensorFlow. In Python Data Analytics; Apress: Berkley, CA, USA, 2018. [Google Scholar]
Name Learning Set | Z1 | Z1 | Z2 |
---|---|---|---|
Model ANN | MLP | RBF | CNN |
Training error | 0.19 | 0.22 | 0.29 |
Validation error | 0.15 | 0.19 | 0.30 |
Testing error | 0.18 | 0.21 | 0.14 |
Quality of learning | 0.88 | 0.87 | 0.87 |
Quality of validation | 0.90 | 0.85 | 0.84 |
Quality of testing | 0.90 | 0.85 | 0.97 |
Learning cases | 520 | 520 | 320 |
Training algorithm | BP50, CG315b | KM, KN, PI | Adam |
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. |
© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
Share and Cite
Przybył, K.; Wawrzyniak, J.; Koszela, K.; Adamski, F.; Gawrysiak-Witulska, M. Application of Deep and Machine Learning Using Image Analysis to Detect Fungal Contamination of Rapeseed. Sensors 2020, 20, 7305. https://doi.org/10.3390/s20247305
Przybył K, Wawrzyniak J, Koszela K, Adamski F, Gawrysiak-Witulska M. Application of Deep and Machine Learning Using Image Analysis to Detect Fungal Contamination of Rapeseed. Sensors. 2020; 20(24):7305. https://doi.org/10.3390/s20247305
Chicago/Turabian StylePrzybył, Krzysztof, Jolanta Wawrzyniak, Krzysztof Koszela, Franciszek Adamski, and Marzena Gawrysiak-Witulska. 2020. "Application of Deep and Machine Learning Using Image Analysis to Detect Fungal Contamination of Rapeseed" Sensors 20, no. 24: 7305. https://doi.org/10.3390/s20247305
APA StylePrzybył, K., Wawrzyniak, J., Koszela, K., Adamski, F., & Gawrysiak-Witulska, M. (2020). Application of Deep and Machine Learning Using Image Analysis to Detect Fungal Contamination of Rapeseed. Sensors, 20(24), 7305. https://doi.org/10.3390/s20247305