India is one of the most important countries that produces cotton as a commercial crop, with an estimated 25% of the population relying on it. An estimation is that 250 million people from all corners of the world are earning their living from the production of cotton. The infections and diseases caused by the fungus, bacteria, and pests affect the production of crops, which shows an impact on the livelihood of the people who are dependent on them. Early and accurate detection of these diseases is essential for reducing the damage to the plant. So, this system is implemented by using a deep-learning-based model to detect the cotton plant leaf disease by using the Multi-Layer Perceptron (MLP) approach. This model can analyze the presence of the disease accurately, which involves the collection of a comprehensive dataset that comprises a large number of images. Preprocessing techniques are applied to enhance the feature standardization and quality of the images. This model is trained to extract the relevant patterns. This system focus on the image detection of the plants and the expansion of the dataset to improve accuracy. Plant disease detection by using deep learning technology will make the process easier when compared to manual processes, and it will also be less expensive. The main goal of the model is to detect the disease of cotton plants at an early stage and prevent further spreading of the disease.