Automatic Determination of the Parameters of Electrical Signals and Functional Responses of Plants Using the Wavelet Transformation Method

Abstract

Smart agriculture management systems with real-time control of plant health and vegetation are recognized as one of the crucial technologies determining agriculture development, playing a fundamental role in reducing yield losses and improving product quality. The earliest plant responses to various adverse factors are propagating stress signals, including electrical ones, and the changes in physiological processes induced by them. Among the latter, photosynthesis is of particular interest due to its key role in the production process. Of practical importance, photosynthesis activity can be registered not only in contact mode but by remote sensing using optical methods. The aim of the present work was to develop the approach to automatic determination of the main parameters of electrical signals and changes in photosynthesis activity and transpiration using continuous wavelet transform (CWT). Applying CWT based on derivatives of the Gaussian function allows accurate determination of the parameters of electrical signals as well as induced physiological responses. Moreover, CWT was applied for spatio-temporal mapping of the photosynthesis response to stress factors in pea leaf. The offered approach allowed automatic identification of the response start time in every pixel and visualization of the change propagation front. The results indicate high potential of CWT for automatic assessment of plants stress, including monitoring of plant health in large-scale agricultural lands and automated fields.

Burning-induced electrical signals influence broadband reflectance indices and water index in pea leaves

“…Measurements of the electrical responses in plants can be used for displaying the action of stressors, which is based on different methods of identification and classification of electrical responses (e.g., machine learning algorithms, interval arithmetic, wavelet analysis, etc. ), [28][29][30][31][32][33] including a systemic analysis of all types of electrical activity as "plant electrome." 34,35 However, measurements of electrical activity are based on the application of electrodes and cannot be used for remote sensing plant systemic response on action of stressors.…”

Multiclass classification of environmental chemical stimuli from unbalanced plant electrophysiological data

“…Water stress in different intensities in the form of high and medium drought and its correlation with plant electrical response has been used for classification in olive trees [ 23 ]. Wavelet transform based analyses have also been reported on plant signals and photosynthetic activities in [ 24 ]. Salt stress effects as a binary classification was attempted using a convolutional neural network with generative adversarial network (GAN) data augmentation for performance boosting [ 25 ].…”