अमूर्त

Optimization of CNN with LSTM for Plant Classification and Prediction

Anuradha Tyagi, Deepika Punj, Shilpa Sethi

Plant diseases are a distinct threat to the field of agriculture and plantation around the world. Diseases that are not properly treated will result in a reduction the harvest of the crop. Early identification of the disease is a very important thing. This is to prevent the disease from spreading to other plants. Detection common plant diseases are made by direct observation of each plant. Whereas, the motive behind the research is the find out the disease at the early stage enabling the conservative management technique to initiate the treatment and avoiding the spreading of disease in remaining plants or leaves. The proposed scheme is automatic detection analogy which based on observation using digital image processing supported by the development of visual technology and digital products. The detection of plant disease is based on the leaves got sick due to insects, bacteria or fungi. However, the scheme inculcates the automatic detection of disease in plant and deficiency occurred due to the complex nature of the disease. Therefore, the proposed scheme identifications of type of disease definitions use a gray level comprising CNN (Convolutional Neural Network) and Long Short Term Memory (LSTM) with efficiency any accuracy.