Detecting spikes of wheat plants using neural networks with Laws texture energy

Table 2 Evaluation of spike detection

Experiment	Total number of images^a	Correct^b	Incorrect^b	Accuracy^c (%)
2013	194	168	26	86.6
2014	206	168	38	81.5

^aFor 2013 dataset, we selected the image on the last imaging day for each plant; for 2014 dataset, in order to get more sample images, we took few images for each plant on the last few imaging days
^bManually check whether all the spikes in the image are detected, if all the spikes are recognized, and no any misclassified, we defined as correct; If there is some spikes are not recognized, or misclassified, we defined as incorrect
^c \(Accuracy = \frac{{N_{correct} }}{{N_{total} }} \times 100\%\)

ISSN: 1746-4811