Design and Implementation of a Measuring Device to Determine the Content of Pigments in Plant Leaves

Abstract

The design and implementation of a measuring device for the determination of pigment content in plant leaves is a topic of essential importance in plant biology, agriculture, and environmental research. The timely and sufficiently accurate determination of the content of these molecules provides valuable insight into the health, photosynthetic activity, and physiological state of plants. This paper presents the key aspects and results of the development and implementation of such a measuring device. It makes it possible to measure a larger number of pigments per type compared with the devices for commercial use that are currently known to us, and the accuracy of measurements depends mostly on the specific type of plant that is being tracked. The developed device presents a measurement accuracy ranging between 72% and 97% compared with a reference method and between 87% and 90% compared with a reference technique. Also, by using the device, a significant reduction in time and required resources can be achieved in measuring the content of pigments and nitrogen in plant leaves. This is a prerequisite for the more effective monitoring of the growth and health of plants, as well as optimizing the process of growing and caring for them. The work will be continued with the focus of the research aimed at generalizing the models for determining pigments and nitrogen in plants.

1. Introduction

Pigments, representing organic molecules, have a decisive role in the process of plant photosynthesis. They absorb different wavelengths of light and convert them into energy. The energy thus obtained from sunlight is transformed into chemical energy, which is used for the synthesis of glucose and other sugars, playing a key role in the production of nutrients. Pigments in plants are found in the chloroplasts of their cells, the most common being chlorophyll, carotenoids, and anthocyanins. Each of them has a specific function and can be an informative sign of plant condition and growth [1].

Methods of analytical chemistry, including chlorophyll extraction and spectrophotometric measurement, are classified as sufficiently accurate for determining the concentration of pigments in plant leaves [2]. Obtaining reliable results with these techniques involves using solvents such as acetone, ethanol, or methanol and calculating extinction coefficients to convert absorbance values into pigment concentrations [3]. However, methods based on non-destructive techniques, such as digital photography, aerial photography, and miniaturized systems for the estimation of pigment concentration, have been developed and researched in recent years. Regardless of these developments, analytical methods remain the gold standard. However, non-destructive techniques have practical advantages, such as the direct availability of results, the fast time it takes to obtain them, and the easy periodic monitoring of plants.

The development of mobile devices to determine the number of pigments in plants is not new. Different manufacturers offer such technical means in commercial networks. The more common mobile devices measure the properties of transmitting light in the VIS and NIR spectral ranges. The light sources are LEDs with a fixed spectral wavelength. The majority of devices are primarily designed to measure chlorophyll. Nitrogen can be measured indirectly based on chlorophyll readings. Some of the devices have a GPS sensor for positioning and memory for the measurement data. Some meters, such as the MPM-100 (Opti-Sciences, Hudson, NY, USA), are stand-alone computing stations capable of displaying spectral data and performing calculations. The N-tester device (Yara International ASA, Oslo, Norway) connects to the camera of a mobile phone and uses all the functionality of this type of communication means.

The mentioned devices offer portability, versatility in pigment measurement, and the utilization of relevant spectral ranges. However, the limitations include a limited range of measurements, fixed wavelength LEDs, potential limitations in functionality and calculations, and dependency on mobile phone connectivity for some devices. These factors should be considered when selecting and utilizing these devices for plant pigment analysis.

The development of mobile devices for the determination of pigments in plants is the subject of scientific developments at the present time [4,5,6]. In recent years, mainly low-cost devices have been created. Devices are proposed for the determination of chlorophyll and flavonoid pigments, as well as nitrogen pigments [7,8,9]. Spectral characteristics in the range 405–1100 nm (VIS and NIR) have been used. The accuracy achieved against a reference measurement is 63–99%. Photosensitive elements, spectrophotometric sensors, and video cameras are used as detectors [10,11,12].

The mentioned devices offer affordability, versatility in measuring different pigments, and the utilization of a broad spectral range. However, limitations include potential variations in accuracy, a lack of detailed performance information, limited information on data processing and analysis capabilities, and potential constraints associated with low-cost designs. These factors should be considered when selecting and utilizing these devices for pigment analysis in plants.

According to Kamarianakis et al. [12], chlorophyll meters provide a convenient tool for plant health assessment, but cost, availability, and device accuracy are important considerations that may limit their widespread use and accessibility. The proposed low-cost device shows promise as an alternative, but further development and validation are necessary to establish its reliability and suitability in various agricultural contexts.

Ibrahim et al. [11] present a promising and accessible approach to estimating chlorophyll content in lettuce leaves using a smartphone-based image acquisition technique. While the method demonstrates significant correlations with chlorophyll content, it is important to consider the limitations and potential variability associated with this approach. Further research and validation are necessary to assess the technique’s performance across different plant species and lighting conditions and enhance its accuracy and consistency.

Urbanovich et al. [13] present a promising application of machine learning, specifically, the random forest algorithm, for predicting chlorophyll content based on the reflection spectra of plant leaves. The study benefits from a diverse sample set and comparative analysis of different models. However, the limited wavelength range, focus on specific plant species, performance variability, and variations in comparison with other models should be taken into account when interpreting and applying the findings.

Barman et al. [14] present a cost-effective solution for predicting chlorophyll content in citrus leaves using smartphone images and predictive models. However, the limited scope, lack of comparison with existing methods, and absence of a detailed evaluation of prediction errors warrant further investigation and validation to assess the method’s robustness and generalizability.

The presented data from the available literature provide a summary of the advantages and limitations of the measurement methods discussed. These methods offer several benefits, including the ability to collect data in situ (on-site) at a lower cost compared with traditional methods, in real time, without causing harm to the plant, potentially with higher accuracy. They offer advantages such as accessibility, profitability, real-time data acquisition, and non-destructive measurements.

However, it is important to acknowledge the limitations associated with these methods. They may exhibit lower precision or accuracy compared with laboratory-based techniques. There could also be limitations in terms of sensitivity or dynamic ranges, which may affect the range of measurable values. Environmental factors, such as changes in lighting conditions or temperature, can influence the measurements obtained. Furthermore, there might be restrictions on the types of pigments that can be effectively measured using these methods. Additionally, the successful utilization of these methods may require specialized training or technical expertise to ensure accurate and reliable results.

While these measurement methods offer several advantages over traditional approaches, it is crucial to consider their limitations and make informed decisions based on the specific requirements and constraints of the research or application at hand.

The aim of this work is to develop and test a measuring device that can be sufficiently accurately and reliably used in determining the pigment content of plant leaves. This will improve the quality of data collected on plant leaf pigment content, which could have important implications for research in a number of fields, including plant biology, agriculture, and the environment.

To achieve the formulated goal, it is necessary to perform the following tasks:

• Design a new measuring device that can accurately determine the pigment content of plant leaves.
• Implement the measuring device and test its performance against existing methods of measuring pigment content in plant leaves.
• Evaluate the reliability and accuracy of the measuring device in comparison with existing methods.

The following hypotheses can be put forward:

• The measuring device will provide sufficiently accurate and reliable measurements of pigment content in plant leaves from existing methods.
• The device will be able to measure a wider range of pigments than existing methods.

The contribution results can be summarized as:

• A measuring device has been developed that has sufficient accuracy and reliability compared with existing methods of determining the pigment content of plant leaves.
• Appropriate regression models were selected for the determination of pigments in plants. They help expand the range of measurement types when assessing the condition of plants.
• A selection of technical means of measurement and management has been developed; these have sufficient accuracy and a low cost and are suitable for performing the tasks required to determine the pigments of plants.

The device was tested and found to provide sufficiently accurate and reliable measurements of pigment content in plant leaves compared with existing methods. This ensures the quality and reliability of the data collected. Unlike previous methods, the proposed device can measure a wider range of pigments. This expanded capability allows for a more comprehensive analysis of plant pigments and provides a more detailed understanding of their function, distribution, and regulation. The device has the potential to save time and resources in the measurement of pigment content in plant leaves. Its efficiency makes it suitable for practical applications such as crop monitoring, where quick and accurate measurements are essential for effective decision making.

This work is organized in the following order: In the “Materials and Methods” section, the research objects are defined. A methodology for reference measurements of chlorophyll is presented. The models for determining basic pigments in plants are described. Measuring and controlling the technical means and methods for their analysis are selected, and the statistical methods used are described. In the “Results” section, we present an electrical diagram, a block diagram of the software, and a graphical user interface of the proposed device. Correction equations for optical sensor data are presented. The results of the literature data analysis are presented. Results from actual plant leaf measurements are also presented. In the discussion part, a comparative analysis is made with accessible literary sources. The results obtained are summarized in the “Conclusion”. Directions for future research are provided.

2. Material and Methods

2.1. Main Measured Characteristics of Plants

The object of this work is to determine the main characteristics of plants related to the following pigments: chlorophyll, carotenoids, anthocyanins. Flavonoid and nitrogen content are also considered.

Table 1. Description of main characteristics of plants considered in this work.

Characteristic	Description
Chlorophylls	Chlorophyll is a green pigment in plants that has a critical role in photosynthesis. Measuring the chlorophyll content of plants can provide insight into their health, growth, and stress responses.
Carotenoids	Carotenoids are pigments that are responsible for the orange, red, and yellow colors of plants. In addition to their role in coloration, carotenoids also have important functions in photosynthesis and plant protection against environmental stressors.
Flavonoids	Flavonoids are a diverse group of plant pigments that have a wide range of functions, including UV protection, pollinator attraction, and defense against pathogens and herbivores. Measuring the flavonoid content of plants can provide insight into their responses to biotic and abiotic stressors.
Anthocyanins	Anthocyanins are water-soluble pigments responsible for the red, purple, and blue colors of plants. In addition to their role in coloration, anthocyanins also have important functions in protecting plants from UV radiation and environmental stressors.
Nitrogen	Nitrogen is an essential nutrient for plant growth and development, playing a critical role in physiological processes such as photosynthesis and protein synthesis. Measuring the nitrogen content of plants can provide insight into their nutritional status and also be used to optimize fertilization practices.

provides a rationale for the main characteristics of plants investigated in this work.

2.2. Reference Measurement of Chlorophyll

Chlorophyll extraction was performed according to the methodology presented by Pompelli et al. [3], with some modifications. A leaf mass weighing 5 g was ground with 2 mL of acetone (80%) at a temperature of 4 °C, in combination with 0.1% sand, in order to prevent chlorophyllase activity. After grinding, the samples were supplemented with 23 mL of acetone. They were left at a temperature of 4 °C for 2 h. Then, they were filtered, and the filtrate was placed in glass cuvettes. A HALO SB-10 UV-VIS-NIR spectrophotometer (Dynamica Scientific Ltd., Livingston, UK) was used, with a measurement range of 200–1200 nm. The absorbance of the filtrates at 664, 649, and 665 nm was determined. The amounts of chlorophyll a and b, µmol/m², were determined using the following formulas:

$$Ch{l}_a=12.19A_{665}-3.45A_{649}$$

(1)

$$Ch{l}_b=21.99A_{649}-5.32A_{664}$$

(2)

The total amount of chlorophyll used in the work was (Chl_a + Chl_b).

2.3. Models for Indirect Measurement of Basic Plant Characteristics

The amounts of chlorophyll, carotenoids, flavonoids, anthocyanins, and nitrogen were calculated. Data from the chlorophyll content index (CCI) [1], determined at specific spectral wavelengths in the VIS and NIR ranges, were used. T931 is transmittance at 931 nm (NIR spectral range), and T653 is transmittance at 653 nm (VIS spectral range).

Chlorophyll content (CC) was determined using the formulas presented by Parry et al. [15]. The carotenoid content (CrC) was determined after converting formulas presented by Bielinis et al. [16] and Shah et al. [17]. Equations for flavonoids were derived from data from Abdallah et al. [18], D’Angiolillo et al. [19], and Zhang et al. [20]. Anthocyanin content (AC) was determined after the conversion of a formula presented by Hlavinka et al. [21]. The nitrogen content (NC) was obtained using a formula presented in [22].

The determination of pigments and nitrogen in plants from data from CCI measurements has the following advantages: the method is non-destructive and allows for quick and convenient correlation with other main characteristics of plants and their growing conditions. Of course, some limitations can also be pointed out, such as indirect evaluations and differences with laboratory measurements of the respective pigments.

$$\text{Chlorophyll Content Index}\hspace{1em}\hspace{1em}CCI=\frac{T931}{T653}, \mathrm{a}.\mathrm{u}.$$

(3)

$$\text{Chlorophyll Content}\hspace{1em}\hspace{1em}CC=-84.3+98.6CC{I}^{0.505}, \mathsf{\mu }\mathrm{mol}/{\mathrm{m}}^2$$

(4)

$$\text{Carotenoid Content}\hspace{1em}\hspace{1em}CrC=-496.53037+449.76517\times CC{I}^{0.215}, \mathrm{mg}/\mathrm{g}$$

(5)

$$\text{Flavonoid Content}\hspace{1em}\hspace{1em}FC=-23.09896+12.67864\times CC{I}^{0.215}, \mathrm{mg}/\mathrm{g}$$

(6)

$$\text{Anthocyanin Content}\hspace{1em}\hspace{1em}AC=-5.28456+2.37404\times CC{I}^{0.215}, \mathrm{mg}/\mathrm{g}$$

(7)

$$\text{Nitrogen Content}\hspace{1em}\hspace{1em}NC=-3.5389+3.33079\times CC{I}^{0.215}, \mathsf{\mu }\mathrm{mol}/{\mathrm{m}}^2$$

(8)

The CCI was calculated according to the technical specifications of a company (Opti-Sciences, Inc., Hudson, NY, USA). Chlorophyll content (CC) was calculated according to a generalized model presented by Parry et al. [15].

2.4. Selection of Measurement and Control Technical Means

The selected technical means for the implementation of the project were chosen in accordance with the goal of achieving sufficient accuracy with the measurements, accessibility in the presented information on the values of the pigments in the plants, reliability and durability with the components in order to be independent of the environmental conditions in which these measurements were carried out, and compatibility and integration based on communication protocols and software interfaces, allowing for interactions with other systems and devices.

Spectrophotometric sensor with a VNIR range (300–1100 nm), TCS230 (Texas Advanced Optoelectronic Solutions Inc., Plano, TX, USA): Its settings were adjusted with a color scale with 24 color fields: Danes-Picta Color Chart BST11 (Danes-Picta, Praha, Czech Republic). Data from the TCS230 sensor setup are presented in Appendix A. The improvement in color accuracy achieved by the correction can be visually evaluated by comparing the corrected RGB values with those of the standard scale. In this way, known deviations or inaccuracies in the sensor measurements are accounted for, providing a more precise representation of the measurement data.

Mega 2560 single-board microcomputer (Kuongshun Electronic Ltd., Shenzhen, China).

Touchscreen display: Nextion Enhanced 2.8” NX3224K028 (ITEAD Intelligent Systems Co., Ltd., Shenzhen, China). The graphic interface of the display was developed in the program Nextion Editor v.1.65.1 (ITEAD Intelligent Systems Co., Ltd., Shenzhen, China).

Cold white LED with a diameter of 5 mm; luminous intensity, 5600 mcd; model VLHW5100 (Vishay Intertechnology Inc., Malvern, PA, USA).

Battery-type Power Bank, 10,000 mAh Supreme 10HD (Hama GmbH & Co. KG, Monheim, Germany).

Junction box 120 × 100 × 68 mm with protection class IP54 (Jv Electric Ltd., Chernozemen, Bulgaria).

2.5. Determination of Transmittance Spectral Characteristics

The algorithm for obtaining spectral pass characteristics is presented in the form of pseudocode in Figure 1. Arrays are declared containing correspondence functions for the red, green, and blue channels and the constant value “c”. Compliance function data for the TCS230 sensor are presented in Appendix B. The arrays are 65 elements long. The variables (rm, gm, and bm) containing the maximum transmittance values for each color channel are set. Also, the measured variables (rc, gc, and bc) from the object are indicated. The maximum spectra (spectra2) were calculated from the data for rm, gm, and bm, and those for the object (spectra1) were calculated from rc, gc, and bc. Transmission spectra were calculated based on the ratio of spectra1 to spectra2.

2.6. Test of the Proposed Equations on Literature Data

Data from the available literature were used [6,15,19,23]. The data are for the following plants: corn, rice, European oak, potatoes, strawberries, tomatoes, barley, wheat, and neotropical trees. The total sample size is n = 640.

2.7. Comparative Analysis of Measurements with the Proposed and Commercial Measuring Devices

A comparative analysis was made with data from measurements taken with the device proposed in this work and the CCM-200 Plus Meter (Opti-Sciences, Tyngsboro, MA, USA). The main characteristic measured with the commercial device was the chlorophyll content index (CCI).

2.8. Statistical Processing of Experimental Data

A linear regression model of a particular type was used to compare data measured using the reference method or device and those measured with the proposed device:

$$y=ax+b$$

(9)

where y is the value measured with the device; x is a value determined using a reference method.

The coefficient of determination, R²; the sum of squared errors (SSE); and the root-mean-squared error (RMSE) were determined. The errors were calculated according to the following formulas:

$$SSE={\displaystyle \sum }_{i=1}^n{\left(y_{i pred}-y_i\right)}^2$$

(10)

$$RMSE=\sqrt{\frac{1}{n}{\displaystyle \sum }_{i=1}^n{\left(y_{i pred}-y_i\right)}^2}$$

(11)

The experimental data were processed with the Matlab 2017b software system (MathWorks Inc., Natick, MA, USA). All data were processed at a significance level of α = 0.05.

2.9. Comparative Analysis of the Obtained Results

The comparative analysis of the results included benchmarking, a procedure for comparing a proposed device with existing devices. This tool allows us to assess whether the results obtained in the study complement or exceed the known performance of devices already on the market. Through this comparative analysis, researchers can determine the extent to which their study results match or improve upon the current state of the art. Benchmarking provides a valuable framework for evaluating the competitiveness and potential market value of the proposed device, which ultimately forms the basis for further research, development, and engineering.

2.10. Determining Uncertainty and Quality Assurance

Table 2. Stages of determining the uncertainty and quality assurance of the proposed device.

Stage	Determining Uncertainty	Quality Assurance
Calibration	Compare device readings against known references	Regular calibration and maintenance
Reproducibility	Measure multiple samples; calculate variability	Standard operating procedures (SOPs)
Inter-laboratory comparisons	Compare device measurements with reference methods	Quality-control samples
Error analysis	Identify and quantify potential sources of errors	Documentation and traceability
Uncertainty calculation	Combine uncertainties from different sources	User feedback and support

presents the stages and procedures for determining the uncertainty and quality assurance of the proposed device. Determining the uncertainty and quality assurance of the proposed device involves assessing the potential variations and errors in its measurements.

The following considerations are taken into account:

• The standard method of measurement involves a spectrophotometer with a cuvette and the standard methodology for the determination of chlorophyll content;
• The proposed device is presented in this work as a new development;
• The compared device is a CCM-200 Plus that Trakia University, Bulgari, owns, and it is used in the areas of ecology and agriculture sciences.

3. Results

3.1. Construction and Setup of a Measuring Device for Determining Plant Pigment Values

Figure 2 shows the distribution of data between the real values of the RGB model for the reference scale and those measured with a TCS230 sensor. The accuracy is sufficient, but the deviations between the real and sensor-measured values of the color components are due to the lack of an infrared filter on the sensor.

The following correction equations are derived:

$$R_c=0.839R_m+45.171$$

(12)

$$G_c=0.7305G_m+62.332$$

(13)

$$B_c=0.7213B_m+38.729$$

(14)

where R_c, G_c, and B_c are the corrected values of the RGB components; R_m, G_m, and B_m are the values measured with the sensor.

In Figure 3, the RGB values are presented in three variants: the original TCS230 sensor measurements, those from a standard scale, and their corrected values after applying the correction equations. When observing the corrected values, it became apparent that the colors noticeably converged with those of the standard reference scale.

Figure 4 depicts an electrical diagram of the developed measuring device. The entire system is powered by a Power Bank-type battery, which provides the necessary electrical supply. The LED is activated by a 200 Ω resistor connected to its anode. The display is connected to the primary serial port of a Mega 2560 microcontroller, utilizing pins D0 and D1 for data output. The control inputs (S0–S3) of the TCS230 sensor, responsible for switching the R, G, and B color matrices, are connected to pins D4–D7 of the microcontroller. The frequency output (OUT) of the sensor is connected to pin D8 of the microcontroller, allowing for data acquisition and processing. This electrical setup enables the proper functioning of the measuring device, facilitating the accurate measurement and display of data.

A block diagram of the algorithm for the operation of the developed device is shown in Figure 5. At the beginning of the single-board computer program, the color-matching functions of the TCS230 sensor are introduced. The control pins of the four sensors, the R, G, B, and C matrices, are defined. The device is initialized, and the communication speed between the single-board computer and the display over the serial channel is set to 9600 bit/s. The maximum transmittance values are determined. When the device is initially stratified, the sensor cell must be tightly closed, and there can be no object between the diode and the sensor. The sensor’s R, G, B, and C data are read sequentially, and correction equations are applied to them. The corrected values are sent to the GUI for visualization. The calculated RGB values are converted into transmission spectra. The ratio between the measured and maximum transmittance spectra is calculated, which represents the final spectral transmittance characteristic. The data from it are sent to the display for visualization. Based on the data obtained from the last spectral characteristic, the CCI is calculated, and from it, the pigment and nitrogen contents of the measured plant are determined.

Figure 6 shows the graphical user interface (GUI) screens of the developed device. The graphical interface consists of four main screens. After turning on the device, the “Main Menu” screen appears, through which the others are accessible. The “Measure” screen presents, in numerical form, the measurements of the main pigments and nitrogen. The “Spectra” section shows the corrected RGB values, the transmission spectra at different spectral wavelengths, and data on the transmission spectral characteristics in graphical form. The “Help” menu displays information about the main measured values from the device, including name and measurement units. All screens have a “Home” button that returns the user to the “Main Menu”.

Figure 7 shows an overview of the developed measuring device. All the main elements are mounted in a box. The display and power button are mounted on its upper side. The sensor part, which includes the TCS230 and the LED diode, extends out of the device with extension cables in order to facilitate the measurement of plant leaves.

The procedure for operating the proposed device is outlined below in Figure 8. This flowchart outlines the step-by-step procedure for operating the proposed device, starting from powering it on to accessing the recorded data and utilizing the device’s features.

By following this procedure, users can effectively utilize the proposed device to measure and record various parameters related to leaves. Firstly, the device is started by pressing the ON/OFF button. It is important to note that no leaf should be placed in the sensor cell during startup. From the main menu, the user navigates to the “Measure” screen. A leaf to be measured is then placed in the device, and after a 5 s interval, the device reads and records data related to chlorophyll, carotene, anthocyanins, and other relevant measurements. The recorded characteristics can be accessed from the RGB and spectral data screens, allowing for further analysis, such as calculating leaf colors and spectral characteristics. Additionally, if needed, the user can refer to the “Help” screen for additional guidance or information.

The procedure described is relatively simple and clear, making it easy for users to understand and use the device. The steps are presented in a logical order, starting with turning on the device and progressing through the different screens for measurement and analysis.

3.2. Test of the Proposed Equations on Literature Data

A comparative analysis was made of literature data for plants, which were compared using the calculated amount of chlorophyll from the CCI according to Equation (3), which was used for the measuring device proposed in this work.

Figure 9 shows the results of this analysis. The relationship between actual and measured chlorophyll content data is described by R² = 0.73, SSE = 9.36, and RMSE = 1.21. The obtained coefficient of determination values and the presented errors of the model prove that it is suitable for predicting chlorophyll values using the CCI. Based on the above, it can be concluded that the equation used has sufficient accuracy and can be used for the device being developed.

3.3. Results Obtained by Testing Leaf Samples

Figure 10 shows the results of the comparative analysis between the developed device and a standard method for the determination of chlorophyll content. When plotting the CCI versus the CC, it can be seen that, for agricultural crops, the data partially overlap, while for the tree species, they differ from them. Also, for woody species, the data are distributed significantly farther from the ideal straight line compared with the other plants considered. This may be due to more outliers and specific features of the plant. In the remaining cases, the plant data are distributed close to a perfectly straight line. The greatest scattering is observed at high values of chlorophyll content.

Table 3. Regression results for the CC, measured with the standard method and proposed device.

	Plant	Beech	Corn	Wheat	Tomato	Barley
Characteristic
R2		0.72	0.94	0.92	0.89	0.97
SSE		7.78	1.53	3.36	4.27	1.31
RMSE		5.37	2.38	3.53	3.98	2.21

shows the results of a comparative analysis of data gathered with the proposed device and a standard method for determining the CC. The lowest values of the coefficient of determination and relatively high error values are observed for the tree species. In the remaining cases, the coefficient of determination is in a range of 0.72–0.97, the SSE is 1.31–4.27, and the RMSE is 2.21–3.98. This is further evidence that the accuracy of determining pigments with a non-contact device depends on the type of plant being measured.

Figure 11 shows data from measurements with the proposed device and commercial (compared) device (CCM-200 plus) for the determination of chlorophyll content. When comparing the data for the CCI, it can be seen that it is evenly distributed around a perfectly straight line. It can be stated that the measurements with the device proposed in this work correspond to up to 90% of those of the commercial one. After calculating the chlorophyll content (CC), it can be seen that, initially, at low chlorophyll values, there is a greater scatter of data around the ideal line. This also leads to a decrease in measurement accuracy for the proposed device.

Table 4. Pigments and nitrogen, calculated based on data from CCM-200 Plus and the proposed device. All data have statistically significant differences at p < 0.05.

Characteristic		Plant	Beech	Corn	Wheat	Tomato	Barley
	Device
CCI	CCM-200 plus		3.25 ± 1.18	8.79 ± 2.27	17.17 ± 2.87	25.74 ± 2.24	32.25 ± 2.2
	Proposed device		3.77 ± 2.53	8.07 ± 4.09	17.19 ± 5.25	31.33 ± 6.2	37.12 ± 4.88
CC	CCM-200 plus		94.4 ± 22.95	211.15 ± 64.83	330.06 ± 83.59	424.06 ± 63.97	485.4 ± 62.41
	Proposed device		108.48 ± 73.15	198.63 ± 116.55	330.3 ± 143.56	477.15 ± 163.38	527.35 ± 135.26
CrC	CCM-200 plus		82.86 ± 30.31	221.13 ± 39.93	332.28 ± 67.67	407.65 ± 38.6	452.59 ± 36.19
	Proposed device		101.87 ± 52.47	208.03 ± 112.4	332.49 ± 146	446.71 ± 169.22	481.73 ± 135.93
FC	CCM-200 plus		6.77 ± 9.96	2.87 ± 7.98	0.27 ± 7.19	2.39 ± 8.01	3.66 ± 8.08
	Proposed device		6.23 ± 7.62	3.24 ± 5.93	0.27 ± 4.99	3.49 ± 4.33	4.48 ± 5.27
AC	CCM-200 plus		2.24 ± 2.83	1.51 ± 2.46	0.92 ± 2.32	0.53 ± 2.47	0.29 ± 2.48
	Proposed device		2.14 ± 2.4	1.58 ± 2.08	0.92 ± 1.9	0.32 ± 1.78	0.14 ± 1.96
NC	CCM-200 plus		0.75 ± 0.09	1.77 ± 0.43	2.6 ± 0.64	3.15 ± 0.42	3.49 ± 0.4
	Proposed device		0.89 ± 0.52	1.68 ± 0.97	2.6 ± 1.22	3.44 ± 1.39	3.7 ± 1.14

CCI—chlorophyll content index; CC—chlorophyll content; CrC—carotenoid content; FC—flavonoid content; AC—anthocyanin content; NC—nitrogen content.

shows the results of the calculation of plant pigments from the CCM-200 Plus device and the device proposed in this work. The proposed device generates pigment values that are close to those of CCM-200 Plus, with relatively small differences observed for most pigment content indices. However, for chlorophyll content (CC) and carotenoid content (CrC), the device tends to show higher values, especially for tomatoes and barley, indicating greater differences between the two devices for these specific pigments and plants.

Table 5. Transmittance-based portable devices for the determination of pigments in plants.

Model	Manufacturer	Spectral Range	CC	CrC	Ant	Fl	NC
SPAD-502+	Konica Minolta Sensing Inc., Osaka, Japan	VIS/NIR	Y	N	N	N	Y
atLeaf CHL PLUS	FT Green LLC, Wilmington, NC, USA	VIS/NIR	Y	N	N	N	N
CCM-200 Plus	Opti-Sciences, Hudson, NY, USA	VIS/NIR	Y	N	N	N	Y
ACM-200 Plus	Opti-Sciences, Hudson, NY, USA	VIS/NIR	N	N	Y	N	N
MPM-100	Opti-Sciences, Hudson, NY, USA	VIS/NIR	Y	N	Y	Y	Y
Force-A Dualex 4 Scientific	DigitalFoodLab, Paris, France	NIR	Y	N	Y	Y	Y
CL-01	Hansatech Instruments Ltd., Norfolk, UK	VIS/NIR	Y	N	N	N	N
MultispeQ V2.0	PhotosynQ Inc., East Lansing, MI, USA	VIS/NIR	Y	N	N	N	N
MC-100	Apogee Instruments Inc., Logan, UT, USA	VIS/NIR	Y	N	N	N	N
N-tester	Yara International ASA, Oslo, Norway	VIS/NIR	Y	N	N	N	Y
Proposed Device	Authors’ development	VIS/NIR	Y	Y	Y	Y	Y

CC—chlorophyll content; CrC—carotenoid content; Ant—anthocyanins; Fl—flavonoids; NC—nitrogen content; Y—yes; N—no.

presents a comparative analysis of the pigments measured with devices from different manufacturers and the device proposed in this work.

Most commercially available pigment-measuring devices are designed to operate within the visible (VIS) and near-infrared (NIR) spectral ranges. These devices primarily focus on measuring chlorophyll content, with the exception of the “ACM-200 Plus” device. However, none of the compared devices in our study had the ability to measure carotenoids, which is a notable advantage of the proposed device. Additionally, only a limited number of commercial devices are capable of measuring nitrogen levels, whereas the proposed device includes this parameter in its measurements. In terms of the measured characteristics, the proposed device complements the capabilities of the “Force-A Dualex 4 Scientific” device, making it a valuable addition to the existing range of pigment-measuring devices.

Important parts of device development and use are uncertainty and quality assurance. The affected aspects related to uncertainty and quality assessment are presented in

Table 6. Determined uncertainty and quality assurance of the proposed device.

Aspect	Description	Value/Ranges
Calibration and Comparison	The device’s optical sensor is compared and calibrated with a standard color scale. A standard spectrophotometric method is used for chlorophyll content measurement. A comparison is made between the proposed device and other devices.	Coefficient of determination: 0.72–0.97
Regular Calibration and Maintenance	Regular calibration and maintenance practices are implemented for optimal working conditions and accurate measurements.
Reproducibility and Precision	Multiple measurements are performed to assess the reproducibility or variability of the results.	Coefficient of variation (CV): 22–35%
Interferences and Error Analysis	Interferences are identified and addressed through error analysis and uncertainty calculations.
Environmental Factors	Environmental factors, including changes in light conditions, temperature, humidity, atmospheric conditions, and leaf properties, are considered potential influences on measurements.
Calibration and Sampling Consistency	Consistent calibration procedures and sampling techniques are maintained to minimize measurement variability and ensure accuracy.
Standard Operating Procedure (SOP) Development	A standard operating procedure (SOP) is developed for correct device usage, including startup procedures, sample handling, and data recording.

.

The device’s optical sensor is compared and calibrated with a standard color scale. Also, a standard spectrophotometric method is used for chlorophyll content measurement, and a comparison between the proposed and compared devices is made. The coefficient of determination is in a range of 0.72–0.97. Regular calibration and device maintenance will keep the proposed device in optimal working order. This helps maintain accurate measurements over time. Also, this allows us to assess the device’s accuracy and identify any discrepancies.

By performing multiple measurements to assess the reproducibility and variability of the results, the obtained coefficient of variation between the standard method and the compared and proposed devices (CV = (mean/SD) × 100) is from 22% to 35%. These low variability values indicate enough precision in the proposed devic.

It is important to consider and address potential interferences to ensure accurate and reliable pigment and nitrogen measurements when using the proposed device. This is essentially part of error analysis and uncertainty calculations.

Environmental factors, including changes in light conditions, temperature, humidity, and atmospheric conditions, and leaf properties, including leaf thickness, surface structure, and moisture content, can impact the absorption and reflection of light, which are crucial for pigment measurements. The presence of substances such as dust, pollutants, or other pigments in plant tissue can interfere with the accurate measurement of specific pigments. Improper calibration or calibration drift over time can affect the accuracy of measurements. Inconsistent sampling techniques or variations in the size, age, and health of plant leaves can introduce variability to measurements.

A clear standard operating procedure (SOP) was developed to outline the correct use of the device, including startup procedures, sample handling, and data recording.

4. Discussion

Based on the results obtained, it can be stated that the implementation of a device for measuring pigment content involves dealing with several challenges. Calibration is critical to establishing a relationship between measured optical signals and actual pigment concentrations. Accurate and representative calibration standards, reference measurements, and calibration-curve-fitting algorithms are required. The device must also account for potential interferences that may affect pigment measurements.

The affordable device proposed in the present work for the determination of essential pigments and nitrogen in plants removes some limitations of such developments [8,9]. Their capabilities do not fully meet the requirements of all researchers or plant breeders because, in most cases, they are limited to measuring only chlorophyll. Some specialized or advanced measurements may still require more complex and expensive equipment, which limits the range of phenotypes that can be assessed.

In addition to its advantages, the proposed device—similar to other portable commercial chlorophyll-measuring devices, such as the CCM-200 Plus—has the same limitations as commercial devices and those presented in the available literature [2,11,12]. This type of meter relies on conversion equations to estimate actual chlorophyll content based on optical measurements. However, one limitation is that, with these devices, the measurements are taken point-by-point from plant leaves, which limits the determination of plant pigment content to small areas.

In terms of cost, as of March 2023, in Bulgaria, the price of a SPAD-502 unit is EUR 2380, and CCM-200 Plus is EUR 2500. The hardware components of the proposed device are priced at EUR 102.

To overcome this limitation, one approach is to make repeated measurements over larger areas and then average the results. This allows for a more comprehensive assessment of leaf chlorophyll content. Furthermore, both the validated experimental device and the commercial chlorophyll meter offer the significant advantage of providing a quick and easy indication of leaf chlorophyll content, especially in field conditions. Here, the advantage of the proposed device is that it allows for the determination of several plant pigments and nitrogen content.

To properly use the proposed device, quality control samples have to be prepared with known properties to validate the device’s performance and verify its accuracy. The procedure of preparing control samples is well known and described in the available literature [1,4,10]. Also, documentation and traceability, user feedback, and support need comprehensive records of calibration, maintenance, and any modifications or repairs made to the device, which can continuously improve the device based on user experiences and requirements.

Another problem that is important to discuss is the generalization of regression models for the determination of pigments in plants. As can be seen from the results presented above, the accuracy of pigment determination depends on the type of plant being measured. Other influencing factors are the geographical region where the plants are grown, seasonal changes, and altitude [15]. It follows that future research in the subject area should be aimed at generalizing the models for determining both the pigments in plants and the nitrogen in them.

5. Conclusions

In this paper, a measuring device was proposed and tested that can be sufficiently accurately and reliably used to determine the pigment and nitrogen content of plant leaves.

A performance test was performed, and the measurement accuracy of the device was evaluated against existing methods and technical means of measuring pigment content in plant leaves. The measuring device was found to provide sufficiently accurate and reliable measurements of pigment content in plant leaves compared with existing methods. It has been shown that the device can measure a wider range of pigments than previous methods.

The measurement accuracy of the proposed device ranges from 72 to 97% compared with the reference method, and compared with the reference, the technical means’ accuracy ranges from 87 to 90%. This accuracy depends on the type of plant being measured.

The proposed measuring device has the potential to save time and resources in the measurement of pigment content in plant leaves, which may have practical applications in areas such as crop monitoring.

The developed measurement device could help improve the quality of data collected on the pigment content of plant leaves, which could positively impact research in a number of fields, including plant biology, agriculture, and environmental science.

The results of this work may provide new knowledge about the physiology of plant pigments, including their function, distribution, and regulation. Also, new methods for measuring pigment content in plant leaves can be developed to be applied to other areas of research and potentially lead to further advances in scientific knowledge.

In summary, modern non-contact methods and means are increasingly important in the quality assurance of various products and processes and are crucial for the safety and reliability of various engineering systems.

This work should be continued with research aimed at generalizing the models for determining both the pigments in plants and the nitrogen in them.