Portable multispectral plant phenotype imaging system

Videometer Lite adopts an LED strobe light source system, effectively combining 7 wavelength measurements and generating a fused spectral image with a unified spectrum, where each pixel corresponds to a different reflection spectrum. This device includes visible light and NIR near-infrared bands for precise and comprehensive detection of crop phenotypes, plant diseases, and more. This portable Videometer Lite can be mounted on a cart stand for use in the field or handheld, making it a multifunctional imaging platform.

Portable multispectral plant phenotype imaging systemMain functions

Combining the advantages of visible light imaging and spectral imaging

Imaging of seeds and disease phenotypes

Portable design, easy to carry to greenhouse or outdoor use

Standard calibration function, data can be repeated

Experienced experts design software based on application experience, which is easy to operate and solves problems encountered in agricultural applications

Built in color correction

Comes standard with 7 spectral bands and is constantly being upgraded

Portable multispectral plant phenotype imaging systemProduct Description

This system can also perform high-throughput imaging measurements on bacteria, fungi, insect eggs, etc. for toxicology or other research, and is used for accurate and comprehensive quality testing of food grains, crops, meat products, and so on. The Videometer system can generate images that can be analyzed using other analysis systems, such as Matlab. Considering that Videometer Lite may need to be frequently taken to greenhouses, fields, or other locations for measurement, it is designed in a portable style.

The working software of VideometerLab Lite is developed by Videometer's powerful bioinformatics and software team, fully considering the needs of practical applications. It is easy to operate and has powerful functions. Videometer is constantly researching and upgrading new algorithms to meet various needs.

The VideometerLab Lite portable seed phenotype multispectral imaging system obtains useful information by measuring the imaging of seeds under 7 different wavelengths (wavelength range 405-850nm) of LED flashes. These images can be analyzed independently or overlaid to create high-resolution color images. Basic integration module, including 7 band multispectral imaging systems. The software can perform color calibration, label recognition, grayscale image conversion, etc.

Application of field multispectral phenotype imaging system

Phenotypic trait analysis/mining, genotype phenotype association

Agricultural breeding

Horticulture, Agricultural Informatics

Fruit quality analysis

Plant Pathology Research

Biomass analysis

Research on Seed Germination

Anti stress research

Directly measured parameters

size

shape

color

Shape and texture

Spectral texture

Spectral components related to surface chemistry

count

Indirect measurement or calculation

Seed purity

Germination percentage

germination rate

Seed vitality

Seed Health

Seed maturity

Seed lifespan, etc

Main Features

Integrated sphere provides uniform and diffuse lighting

Realize spectral imaging and quantitative analysis within 10-15 seconds

7 different wavelengths/light sources

3 million pixels/wavelength, provided, with a resolution of 21 million pixels/frame

Standard equipment includes easy-to-use device calibration

Compared with traditional RGB technology, it has * color measurement function

Automatically switch dynamic range according to application requirements

Long lifespan of light source, up to 100000 hours

Enhanced stability of LED light source technology

Powerful exploration software for research

Easy to use conventional application formula construction tool (modeling)

imaging characteristics

Fast and non-destructive testing

Including processing, each sample processing only takes 10-20 seconds

Combined with other destructive technologies

High flexibility measurement

Main focus: repeatability, traceability, durability, and transferability

Technical Specifications

Complete analysis time 10-15 seconds per sample

Power supply: 5 V DC 3 A

Power consumption 300 VA

Environmental temperature operation: 5-40 ℃, storage -5-50 ℃

Environmental humidity 20-90% RH relative humidity, non condensing

Software alternative: Image Processing Toolkit (IPT)

Spectral Imaging Tool Box (MSI)

Spot Tool Box

Equipment size: 270 mm (h) * 240 mm (w) * 200 mm (d)

Weight: 1.1kg

Case Application

Distinguishing seeds by chlorophyll/maturity

Scientists from the UK are focusing on evaluating advanced imaging techniques for fungal detection and precise quantification of root colonization, and assessing their impact on aboveground health by measuring photosynthetic parameters. The VideometerLab multispectral imaging system was used in the study.

The image shows wheat seedlings infected with 'Take all'. On the left is the original image, with a red arrow indicating the "take all" loss, manually graded; The right image shows the same image analyzed by VideometerLab, which classifies root tissues as diseased (blue) and healthy (orange/yellow).

Imaging of Quinoa downy mildew using Videometer multispectral imaging system

Chenopodium quinoa is a nutrient rich crop widely grown in multiple countries. Fungal diseases such as downy mildew limit grain yield, and cultivating resistant strains, such as those resistant to downy mildew, is the central goal of quinoa breeding.

It is difficult to measure the phenotypic response of quinoa to downy mildew using conventional RGB imaging, as different quinoa genotypes have different green and red spots on the leaves that interfere, as shown in Figures 1 and 2.

Develop image analysis protocols to distinguish healthy quinoa leaf tissue from quinoa leaf tissue infected with downy mildew. Study the severity phenotype and spore formation using Videometer multispectral imaging system.

The severity is the percentage of the area of frontal damage to the entire leaf area. Depending on genotype, the color can be orange, yellow, or red.

Spore formation is the amount of spores above the damaged area, measured as a percentage, and evaluated by measuring the front of the leaf.

Figure 1 Symptoms of Frontal Severity of Leaves

Figure 2: Formation of Spores on the Frontal Surface of Leaves

Multispectral image analysis

The researchers used the VideometerLab 4 multispectral imaging system for multispectral imaging, with an integrating sphere ensuring uniform illumination of the sample (Figure 3). Each acquired image layer consists of 19 different image bands, covering wavelengths from 365nm (UVA) to 970nm (NIR). The resolution of each pixel in the image is~41 µ m. The resolution of each image layer is 2192X2192 pixels.

Image Analysis Severity Model

From the front of the G9 genotype leaves (Figure 4), yellowing phenomenon (A) can be clearly seen, and RGB images (conventional camera, visible light band to the human eye) were taken. (B) and (C) show two bands in the multispectral layer, blue light 490nm (B) and yellow light 570nm (C). Initial labeling was performed on healthy plant tissues and yellowing definitions, and a model (D) was established through transformation. The 19 band information (multiple layers in the image) was transformed into representative pixel range values for the entire layer using nCDA (Normalized Canonical Discriminant Analysis). Afterwards, cutting (E and F) can be used for quantitative analysis of the percentage of yellowed tissue (E yellow) in all images - all strains and genotypes. The specific leaf proportion is 68.0%, or includes red covered spore areas (F), accounting for 18.9%, yellowing (yellow) proportion is 68%, and the combined area of spores and yellowing areas accounts for 75.8%.

Image analysis of spore formation

On the front (bottom) of the leaf, the G9 genotype in the RGB image clearly shows spore formation images (bottom A and B enlarged). Although it is difficult to detect a single band in the visible light range, the blue light band (490nm) is specifically marked here (C). Entering the NIR (780nm) band (with D and E magnification in the lower left), spores were clearly visible. Using this information (only identifying black and gray spores) can help us distinguish the cut spore pixels (F) and quantify the area. The proportion of spores in this leaf is 12.5% (shown in yellow), excluding the yellowing area.

In addition, the spore identification here is more conservative compared to frontal image analysis. The pixel portion of the non black gray area covered (with larger pixels than a single spore) is estimated to have a spore proportion of~23% (not shown here).

image4 (A) sRGB image. (B) Quantitative segmentation of two types: 490nm (blue light), (C), 570nm (yellow), (D) conversion, (E) and (F).

Figure 5 (A) sRGB image, (B) 490nm (blue light), (C) 570nm (yellow), (D) conversion, (E) quantitative segmentation.

result

Figure 6: Distribution of average severity (%) for 133 genotypes

Table 1 Interaction of Quinoa downy mildew by manual and multispectral phenotype imaging