Goal
With the help of a Raspberry Pi and some knowledge of Normalized Differentiation Vegetation Index (NDVI), I was able to code the expression of plants. I seek to create a more just and sustainable world. When you give people the power to make informed choices we can prevent injustices.
Introduction/Background
With the help of cameras using special filters, we can measure a plant's health. Plants are green because they contain a chemical called chlorophyll. Chlorophyll helps them use light from the sun to turn carbon dioxide and water into carbohydrates (this process is known as photosynthesis).
There are many forms of light, many of them not visible to the human eye. A plant can only use certain forms, and one it can’t is green light, and so it gets reflected away making the plants look green.
Plants don’t really like infrared radiation that much because it heats them up if they receive too much. Two types of infrared light, shortwave and longwave. Plants that are healthy will reflect a lot of near infrared light. On the other hand if a plant is dying it will absorb a lot of infrared light.
Filters allow us to prevent certain wavelengths of light from reaching a camera’s sensor. Digital cameras often have infrared filters to stop infrared light from reaching its sensor. In order to detect the amount of infrared light being reflected by plants, we need to use a camera without an infrared filter.
NDVI requires that we increase the contrast in our images
Our images are just a big list of numbers, and we can find the biggest and smallest of those numbers with numpy
Once we have a high contrast image, we can do NDVI calculations
- Taking blue pixels and making them brighter
- Making red pixels darker
- Leaves an image black and white
- Brightest pixels in image indicate healthy plants
- Darkest pixels indicate unhealthy plants, or absence of plants
"The challenges we face are perfectly clear. Doing nothing will cost us everything." - OnePointOne