Instrumentation & Measurement Magazine 24-2 - 99

Table 2 - Classification testing accuracy
Method

Accuracy

Linear Discriminant

54.88

Linear Support Vector Machine

55.33

Ensemble Bagged Trees

53.97

k-Nearest Neighbor

56.01

Binary classification decision tree

55.1

NN

78.26
97.97

CNN

[5]	 K. A. Steen, P. Christiansen, H. Karstoft and R. N. Jørgensen,
" Using deep learning to challenge safety standard for highly
autonomous machines in agriculture, " J. Imaging, 2016.
[6]	 Y. Li, M. Iida, T. Suyama, M. Suguri and R. Masuda,
" Implementation of deep-learning algorithm for obstacle
detection and collision avoidance for robotic harvester, "
Computers and Electronics in Agriculture, vol. 174, Jul. 2020.
[7]	 Z. Zhang, H. Liu, Z. Meng and J. Chen, " Deep learning-based
automatic recognition network of agricultural machinery images,
Computers and Electronics in Agriculture, vol. 166, Nov. 2019.
[8]	 A. Paszke, S. Gross, S. Chintala et al., " PyTorch: an imperative
style, high-performance deep learning library, " Advances in Neural
Infor. Processing Sys., 2019.

one second mark using the same equipment and software. Fig.
5b shows the steps.
For comparison purposes, using the same six features and
same number of samples, five different classifiers yielded the
accuracies given in Table 2.
We can see how promising the use of DL can be, and
we also highlighted the impact that data acquisition and
data preprocessing can have for a system. The topics discussed here not only apply to agricultural machinery, but
these ideas can also play an important role in other DL
applications.

[9]	 M. Abadi, P. Barham, J. Chen et al., " TensorFlow: a system for
large-scale machine learning, " in Proc. 12th USENIX Symp.
Operating Sys. Design and Implementation, 2016.
[10]	MATLAB 9.8 and Statistics and Machine Learning Toolbox 12.0,
The MathWorks, Inc., Natick, Massachusetts, United States.
[11]	 " ImageNet, " ImageNet. [Online]. Available: http://www.imagenet.org.
[12]	P. L. Olson and M. Sivak, " Perception-response time to
unexpected roadway hazards, " Human Factors: The J. Human
Factors and Ergonomics Soc., 1986.
[13]	T. J. Triggs and W. G. Harris, " Reaction time of drivers to road
stimuli, " Monash University Human Factors Group, Report HFR-

Conclusions

12, Monash University, Melbourne, Australia, 1982.

In this paper, the recent techniques for image and signal
classification are reviewed with particular emphasis on agricultural applications. The experiments presented in this
paper highlight the limitation of classical machine learning
techniques versus neural network approaches. As demonstrated, classical machine learning approaches achieve lower
performances compared to neural network and CNN architectures. However, the need for a larger amount of training
data labeled by an expert may not always be available in real
world technologies, and the hardware requirements (GPUs)
for computation may limit the use of such systems in realtime applications.

[14]	A. Simundsson, " A neural network to classify auditory signals
for use in autonomous harvester control systems, " M.Sc. degree
thesis, The University of Manitoba, Canada, 2019.
[15]	J. Ubbens, M. Cieslak, P. Prusinkiewicz, C. Mikolaj and I.
Stavness, " The use of plant models in deep learning: an
application to leaf counting in rosette plants, " Plant Methods, issue
6, 2018.
[16]	R. Girshick, J. Donahue, T. Darrell and J. Malik, " Rich feature
hierarchies for accurate object detection and semantic
segmentation, " Computer Vision-ECCV, pp. 345-360, 2014.
[17]	R. Girschick, " Fast R-CNN, " in Proc. IEEE Int. Conf. Computer
Vision, pp. 1440-1448, 2015.
[18]	S. Ren, K. He, R. Girshick, and J. Sun, " Faster R-CNN: towards

Acknowledgment

real-time object detection with region proposal networks, "

This work was partially supported by the MICINN Grant
RTI2018-095232-B-C21, 2017 SGR 1742.

Computer Vision and Pattern Recognition, 2016.
[19]	D. I. Patricio and R. Rieder, " Computer vision and artificial
intelligence in precision agriculture for grain crops: a systematic

References

review, " Computers and Electronics in Agriculture, vol. 153, pp.

[1]	 A. Sagan, " Agriculture could add $11B a year to Canada's GDP by
2030, " Canadian Business, Aug. 2019.

69-81, Oct. 2018.
[20]	F. Gao, L. Fu, X. Zhang, Y. Majeed, R. Li, M. Karkee and Q. Zhang,

[2]	 F. L. Pryor, " The invention of the plow, " Comparative Studies in
Society and History, vol. 27, no. 4, pp. 727-743, 1985.
[3]	 S. Fountas, C. G. Sorensen, Z. Tsiropoulos, C. Cavalaris, V. Liakos,
and T. Gemtos, " Farm machinery management information

" Multi-class fruit-on-plant detection for apple in SNAP system
using Faster R-CNN, " Computers and Electronics in Agriculture,
vol. 176, 2020.
[21]	P. Christiansen, L. N. Nielsen, K. A. Steen, R. N. Jørgensen and

system, " Computers and Electronics in Agriculture, pp. 131-138,

H. Karstoft, " Deep anomaly: combining background subtraction

2015.

and deep learning for detecting obstacles and anomalies in an

[4]	 M. Islam, Anh Dinh, K. Wahid and P. Bhowmik, " Detection of
potato diseases using image segmentation and multiclass support

agricultural field, " Sensors, vol. 16, 2016.
[22]	J. Burriel-Valencia, R. Puche-Panadero, J. Martinez-Roman, A.

vector machine, " in Proc. IEEE 30th Canadian Conf. Elec. Comp.

Sapena-Bano and M. Pineda-Sanchez, " Short-frequency Fourier

Eng. (CCECE), pp. 1-4, 2017.

transform for fault diagnosis of induction machines working in

April 2021	

IEEE Instrumentation & Measurement Magazine	99
http://www.imagenet.org http://www.imagenet.org
Instrumentation & Measurement Magazine 24-2

Table of Contents for the Digital Edition of Instrumentation & Measurement Magazine 24-2
