How to Easily Detect and Monitor Plant Health

Abstract

This application note explores the use of a Canon RGBNIR camera for plant health inspections. As compared to existing solutions where users combine RGB and NIR cameras for plant health inspections, the Canon RGBNIR is able to provide both RGB and NIR images. This makes inspections using the Normalized Difference Vegetation Index (NDVI) easier for users, as they would not need to calibrate multiple cameras. 

Problem Statement

Over the years, agriculture inspections have become a rising concern in most countries in order to provide safe and healthy foods for all. This inspection applies to food crops and livestock, and if they carry unwanted bacteria or viruses that might harm consumers. However, inspecting plant health conditions may be more challenging than inspecting livestock, since they can not be detected with a human eye or colour camera. This forces most agriculture researchers to seek a solution or suitable hardware to conduct observations on plant health. 

Hence, which hardware would be applicable for this problem?

Figure 1: Setup image with Canon U3-A0500CIR-60 camera and Computar M1628-MPW3, along with some vegetables for plant health inspection.

Solution

For this application, our tech team used a Canon U3-A0500CIR-60 camera with a Computar M1628-MPW3 in their set-up to detect the plant health conditions in everyday vegetables. 

The Canon U3-A0500CIR-60 camera is able to provide both RGB and NIR images, making this camera suitable for monitoring the Normalized Difference Vegetation Index (NDVI) in plants. With the Canon camera, combined with algorithms, users can get images that show the plant’s NDVI range.

In the images below, plant leaves that fall under the red to orange range are in a healthy state, whereas those that fall in the green to blue range are unhealthy, infected by diseases, or withered. With the help of NDVI mapping, agricultural researchers are able to detect problems sooner, taking action to improve plant health sooner before it becomes detrimental to their yields.

Figure 2: Left: Image of harvested vegetables with Canon U3-A0500CIR-60 camera and Computar M1628-MPW3
Right: Scan of harvested vegetables with NDVI mapping

Figure 3: Left: Image of potted plants with Canon U3-A0500CIR-60 camera and Computar M1628-MPW3
Right: Scan of potted plants with NDVI mapping

Figure 4: Left: Image of a tree with Canon U3-A0500CIR-60 camera and Computar M1628-MPW3
Right: Scan of a tree with NDVI mapping

The Canon U3-A0500CIR-60 camera is more suitable for this application as compared to combining an RGB camera with a NIR camera. Setting up the RGB and NIR camera is a time-consuming process for customers, as calibration for the field of view for both cameras is needed. This is to get the same view of both images for the NDVI colour mapping. If the field of view for both cameras is different, the NDVI will show the wrong result, leading to late prevention or action taken to improve the plant. By using a Canon U3-A0500CIR-60 (an RGBNIR camera), users are able to eliminate the process of calibrating the field of view of two images. 

This application set-up is also suitable for monitoring different types of plants, which include trees, crops, and garden plants.

 

Reach out to us!

Find out more about how the Canon RGBNIR cameras can be used for other inspections in different industries!