Introduction: Plants are a significant source of energy. Crop protection and production can be increased using the early and accurate diagnosis of plant diseases. In the old-fashioned environment, the identification is processed whether the plant leaf is healthy or infected either by visual observation or by testing leaves in the laboratory. The visual identification is done by the experts of the plant domain but opinion may vary from expert to expert. Testing of the plant leaf in the laboratory is a very time consuming and strenuous process and hence results may not come on time. Aims: The aim of this research article indicates that the proposed Convolutional neural networks (CNN) provide a healthier solu tion in disease control for capsicum leaf with high accuracy of validation and a faster convergence rate. Methodology: To overcome these issues, image-based plant diseases classification and detection using Convolutional neural networks (CNN) have been presented in the literature. The authors have focused on the capsicum plant (Bell pepper) for this purpose, which belongs to the Grossum cultivar group of the species Capsicum annuum disease. Results: After model development and fitting, the operational performance and quality can be evaluated on the unseen testing dataset. The performance is measured in terms of accuracy. The model accuracy of each block VGG model can be calculated by increasing the convolutional layer and pooling layer. The model accuracy is improved from 84% to approx. 96%. Conclusion: Convolutional neural network is performed to detect, identify and classify the capsicum plant disease in this re search. This research article reconnoitred three different improvements to the baseline model. The performance of the different results can be summarized in the terms of model accuracy.