Prediction of Chlorophyll Content in Different Light Areas of Apple Tree Canopies based on the Color Characteristics of 3D Reconstruction

Abstract

Improving the speed and accuracy of chlorophyll (Ch1) content prediction in different light areas of apple trees is a central priority for understanding the growth response to light intensity and in turn increasing the primary production of apples. In vitro assessment by wet chemical extraction is the standard method for leaf chlorophyll determination. This measurement is expensive, laborious, and time-consuming. Over the years, alternative methods-both rapid and nondestructive-were explored, and many vegetation indices (VIs) were developed to retrieve Ch1 content at the canopy level from meter-to decameter-scale reflectance observations, which have lower accuracy due to the possible confounding influence of the canopy structure. Thus, the spatially continuous distribution of Ch1 content in different light areas within an apple tree canopy remains unresolved. Therefore, the objective of this study is to develop methods for Ch1 content estimation in areas of different light intensity by using 3D models with color characteristics acquired by a 3D laser scanner with centimeter spatial resolution. Firstly, to research relative light intensity (RLI), canopies were scanned with a FARO Focus3D 120 laser scanner on a calm day without strong light intensity and then divided into 180 cube units for each canopy according to actual division methods in three-dimensional spaces based on distance information. Meanwhile, four different types of RLI were defined as 0-30%, 30-60%, 60-85%, and 85-100%, respectively, according to the actual division method for tree canopies. Secondly, Ch1 content in the 180 cubic units of each apple tree was measured by a leaf chlorophyll meter (soil and plant analyzer development, SPAD). Then, color characteristics were extracted from each cubic area of the 3D model and calculated by two color variables, which could be regarded as effective indicators of Ch1 content in field crop areas. Finally, to address the complexity and fuzziness of relationships between the color characteristics and Ch1 content of apple tree canopies (which could not be expressed by an accurate mathematical model), a three-layer artificial neural network (ANN) was constructed as a predictive model to find Ch1 content in different light areas in apple tree canopies. The results indicated that the mean highest and mean lowest value of Ch1 content distributed in 60-85% and 0-30% of RLI areas, respectively, and that there was no significant difference between adjacent RLI areas. Additionally, color characteristics changed regularly as the RLI rose within canopies. Moreover, the prediction of Ch1 content was strongly correlated with those of actual measurements (R = 0.9755) by the SPAD leaf chlorophyll meter. In summary, the color characteristics in 3D apple tree canopies combined with ANN technology could be used as a potential rapid technique for predicting Ch1 content in different areas of light in apple tree canopies.

Editorial for the Special Issue “Estimation of Crop Phenotyping Traits using Unmanned Ground Vehicle and Unmanned Aerial Vehicle Imagery”

“…The methods are broadly classified into two types, and each was applied to the two above mentioned types of phenotyping traits. For biochemical traits, machine learning and artificial intelligence methods, including principal component analysis (PCA), partial least squares regression (PLSR), random forest regression (RF), artificial neural network (ANN), etc., were usually preferred and recommended [10,[14][15][16][17][18]21,23,24,32]. Those methods proved to be very efficient for the various target phenotyping traits.…”

“…For ground platform, ground fixed scanning system [15,22,30,35], handheld-based field measuring [11,14,16,32], mobile ground platform (MGP) [14,20], and lifting hoist-based elevated platform [12,36] were reported for different crops types ( Figure 2). These platforms were easy-to-use with low cost, but data acquisition was semi-automatic.…”

Estimating 3D Chlorophyll Content Distribution of Trees Using an Image Fusion Method Between 2D Camera and 3D Portable Scanning Lidar

“…Such applications have been reported in several studies. A portable scanning lidar that can provide a colored point cloud was used for estimating the chlorophyll contents within trees [16]. Chlorophyll (abbreviated as Chl) and nitrogen contents were also estimated by a portable scanning lidar with a green laser source [17,18].…”

“…To apply the fusion method to various types and sizes of plants, a method is required in which the 2D image is directly fused to the 3D point cloud image without conversion to the polygonal surface mesh. In the studies by [16,26], fusion systems were designed and assembled for specific 2D cameras and lidars. To effectively obtain plant physiological and biochemical information together with the structural features, the fusion method should be easily applicable to different types of 2D cameras and lidars.…”

High-Throughput Phenotyping Analysis of Potted Soybean Plants Using Colorized Depth Images Based on A Proximal Platform

“…A terrestrial laser scanner is a remote-sensing device used to measure the distance from sensors to an object of interest based on the ToF principle. This technique has gained increasing attention in precision agriculture and forestry applications due to its high accuracy and reading speed rate [34]. Although this laser technique exhibits a very high resolution, it also presents disadvantages.…”