Plants are an essential part of the biodiversity and are the key component in the life cycle of nature. to identify different leaves, the performance of Convolutional Neural Network is evaluated by analysing K – fold and mean activation values on Swedish Leaf Dataset. K –fold cross validation allows to maximize the ability of the performance of any neural network . This project aim to showcase that CNNs with K –fold provide impressive accuracy results over the test and validation data with training over Swedish leaf dataset.