Grading rice quality using AI
Raitong, the largest organic rice supplier in Thailand needed an automated way to grade the rice grain to do the following:
- Determine the payment to farmers based on the quality of grain.
- Adjust the speed of the drums during dehusking (faster speed will damage the grain).
- Check the quality of rice for quality control during export and remove any chalky grain.
An automated process that can help to solve the above problem faster and easier using app-based computer vision to detect the length and shape of the grain.
By using an Android app-based computer vision, we detected the length of the rice and determine if the rice grain was whole or broken.This information was also made easily available to the users (Bryan, the farm mill, and the farmers) and a jig was created to separate the grain easily and in around one minute. This innovation presented a 29 minute saving for every 100 grams of rice, and fewer wasted grains during the process of husking.