Low-cost 3D systems: suitable tools for plant phenotyping.

Area of Science:

• Agricultural Science
• Computer Vision
• Plant Biology

Background:

• 3D imaging of plant geometry is crucial for phenotyping and plant breeding.
• Traditional methods like multi-image 3D reconstruction and laser scanning have limitations such as extensive post-processing or high costs.
• There is a need for cost-effective and accurate 3D imaging solutions in plant science.

Purpose of the Study:

• To compare the 3D measuring accuracy of two low-cost 3D imaging systems (David laser scanner, Microsoft Kinect) against a high-precision laser scanner.
• To evaluate the reliability of these low-cost systems for extracting key plant parameters from sugar beet taproots, leaves, and wheat ears.
• To assess the potential of low-cost sensors for automated plant phenotyping applications.

Main Methods:

• Utilized a high-precision close-up laser scanner as the reference method.
• Employed the David laser scanning system and Microsoft Kinect Device as the low-cost 3D imaging systems.
• Estimated 3D measuring accuracy by analyzing deviations on test specimens and evaluating parameters from plant structures (sugar beet taproots, leaves, wheat ears).

Main Results:

• Both low-cost 3D imaging systems demonstrated high reliability for plant phenotyping applications.
• The evaluated parameters extracted from plant geometries showed good accuracy and correlation with reference measurements.
• The study confirmed the suitability of carefully selected low-cost sensors for plant geometry analysis.

Conclusions:

• Low-cost 3D imaging devices are highly reliable for plant phenotyping, meeting the demands of current research.
• These systems offer a cost-effective alternative to expensive laser scanners, with potential for automated applications.
• The findings support the implementation of affordable 3D imaging technologies in plant breeding and phenotyping research.