Computer Vision based grain quality estimation solution in AgriTech to simplyfy grain segmentation & reduce go to market time
The use case covers the assessment of grain quality by performing segmentation and classification on a heap of grains.
To assess the quality of an agriculture yield there is a need for manual labor where a human must take a look at the overall yield to find the ratio of healthy against the total number of grains where grains could be damaged, broken, foreign matter, etc.
With the help of Image segmentation, the model can extract individual grains from a heap and then classify each grain based on the classes provided.
By taking out a small heap from a sack, and then capturing a picture from the mobile, the platform will get the required healthy grain ratio as well as provide details for each type of grain.
Reduces the time taken to segregate the grains based on quality.
The assessed data can be logged in some datasets to gain insights over different timespans, such as months, seasons, etc to make better business decisions.
The platform can be completely cloud-native. The ML model can be offered as an API service to the end-user and can be integrated into any device.