LeafSpec

 

LeafSpec LLC is the ONLY company with the exclusive license from Purdue OTC to commercialize Purdue's handheld leaf imaging technologies.

LeafSpec LLC, Purdue University, and Dr. Jin's lab do NOT have any current relationship with the external company Leaf Tech Ag, and we are NOT responsible for any claims they made about their product

Hyperspectral imaging (HSI) has been helping farmers and researchers produce better-quality food. Unlike traditional imaging techniques, it helps people understand plants better by capturing images in more than 500 colors across visible and invisible ranges. However, many (HSI) systems are large and expensive, and their data suffer from noises of sun angles, leaf angles, and shadows. In 2019, Purdue ABE Sensor Lab developed a handheld hyperspectral imager, LeafSpec. LeafSpec is an easy-to- use and low-cost crop phenotyping sensor with improved measurement accuracy, which could benefit more people in plant science research and agriculture production. For the first time, people can obtain real-time hyperspectral data without the noises.

 

The hardware of LeafSpec was comprised of a push-broom hyperspectral camera (HSC), a leaf scanner with an encoder system for leaf position information, a lightbox as an intensive and uniform beam lighting source, and an ARM-based microcontroller. In each scanning, a smooth and clear hyperspectral image of the entire leaf was obtained by quickly sliding LeafSpec across the leaf from the beginning to the tip. The collected data can provide a resolution of up to 0.4 mm/pixel. Each measurement was geo-referenced by sending processed data to a smartphone and combining it with the GPS location and time information before uploading it to a Geography Information System (GIS) with Digital Ag Map Services and an internet connection.  LeafSpec has various versions to meet different needs. The corn version is designed to scan plants with long leaves, such as corn and sorghum; the dicot version collects data from plants such as soybean, potato and tomato; the wheat version is designed for plants with thin leaves, such as rice and wheat.

 

Purdue researchers and collaborators have been using LeafSpec worldwide to conduct plant studies with high-quality hyperspectral data. The high signal-over-noise feature and high resolution of LeafSpec make it possible for users to detect and classify diseases, nutrient deficiencies, and chemical damages earlier. For farmers, early response means less yield loss and less treatment cost. For companies, early detection significantly reduces the cycle time of product testing/validation. For researchers, our cutting-edge data quality helps them understand plants in a new way. More variations of LeafSpec are under development to provide plant data with better quality and higher throughput. A preview of other variation can be viewed here.