Timely disease detection and management can help optimize crop yield. Corn leaf diseases, if left untreated, can lead to reduced photosynthetic capacity, premature leaf senescence, and overall plant stress, resulting in lower yields. By detecting and addressing diseases early, farmers can mitigate these negative impacts and maintain or improve crop productivity. Relevant features are necessary for machine learning models to learn from. Features can be taken from the pre-processed photos in the identification of maize leaf disease. These features could be more complex ones extracted using convolutional neural networks (CNNs), such as color histograms, texture descriptors, shape-based features, etc. CNNs primarily operate on local image patches and may lack holistic contextual understanding. CNNs may find it difficult to grasp long-range dependencies and relationships throughout the entire image for jobs that call for global context, like comprehending how various portions of a corn leaf interact to identify illness. Transfer learning offers significant advantages in terms of improved generalization, reduced training time, and handling data limitations. By transferring knowledge from pre-trained models to the target task, transfer learning can enhance the performance and efficiency of models, making it a valuable technique for corn leaf disease detection and various other machine learning applications. The proposed uses VGG-16 transfer learning approach to perform multi classification on corn leaves by reducing the losses due to the imbalance datasets.