Coconut tree is one of the most beneficial and popular crops in the world due to its numerous uses. In countries like India, Indonesia, Philippines, Sri Lanka, etc., people primarily depend on coconut farming for their livelihood. Pest infestation or disease can largely affect the production of crops and lead to losses. Finding the disease at its earliest stage of development is necessary for disease control in plants. Image processing technologies, computer vision and deep learning models are now-a-days used to detect disease in crops. Many researchers have developed effective methods using deep learning models for disease detection, Convolutional Neural Networks (CNNs) and various architectures of CNNs are used for Disease Detection in plants. This study aims to review various research methodologies and apply them to a new dataset that includes images of leaves classified primarily into 5 diseases: yellowing, drying, flaccidity, caterpillars, and leaflets. CNN, DenseNet121, InceptionResNetV2, and ANN were major algorithms used in study.InceptionResNetV2 CNN, and DenseNet121 all performed remarkably well and provided the desired accuracy of around 99% and higher.