Crop diseases significantly impact agricultural production, causing substantial losses in quantity and quality. The early detection of diseases through regular monitoring is vital for minimizing these losses. Integrating advanced technologies, such as image processing, machine learning (ML), and deep learning (DL), has led to the development of innovative agriculture solutions. Researchers have utilized these techniques to create automatic crop disease identification systems, offering an innovative approach to address the challenges associated with crop health. The paper used different lightweight convolutional neural networks (CNNs) with modifications. These are good and efficient for classifying images specifically for identifying crop diseases by analyzing crop-leaf images.This cutting-edge approach promises to provide an effective and cost-effective solution for automatically categorizing crop diseases. The performance of the model is assessed using a comprehensive evaluation framework applied to the CCMT dataset that collectively encompasses various crop varieties. The dataset consists of 22 classes and 24,881 rows of images of four different types of crops. The experimental results highlight the superiority of the model over recent DL approaches in the field of crop disease, achieving an impressive accuracy of 99.73% on the VGG16 model. This research contributes to advancing automated crop pest disease categorization, offering a promising solution for practical and flexible agriculture.