A tensorflow model trained to identify weed types in sugarcane farming. Tested and developed using the Tesnorflow library and a custom basic layer augmented with the Inceptionv3 the model consistently achieved identification accuracies north of 92% using the datasets of 5 main sugarcane weeds, mainly:
- Ageratum conyzoides
- cynodon dactylon
- sorghum halepense
- sugarbeat capsella
- sugarbeat galium