Abstract:
Plant disease is a persistent problem for smallholder farmers, threatening
their income and food security. The current surge in smartphone adoption and
computer vision models has presented an opportunity for picture categorization in
agriculture. Convolutional Neural Networks (CNNs) are regarded cutting-edge in
image recognition, with the capacity to produce rapid and precise results.
The efficacy of a pre-trained ResNet34 model in identifying crop disease is examined
in this Project. The proposed model, which is implemented as a mobile application, is
capable of distinguishing several plant illnesses from healthy leaftissue. A dataset is
created in a controlled environment for training and evaluating the model. The
validation findings suggest that the suggested approach is capable of achieving
accuracy. This illustrates the technological viability of CNNs in diagnosing plant
illnesses and paves the way for AI solutions for small-scale farmer
