Sericulture has been the backbone of India’s socioeconomic, economic, cultural, and political developments. Being the second largest producer of silk in the world, sericulture plays a very important role in developing the country, especially with the multiplication of silkworms that produce silk. Silkworm growth is sensitive to temperature and humidity, especially during the larval stage. This paper utilizes webcam technology and image processing to detect silkworms infected with diseases or showing developmental abnormalities. Colour variation patterns in silkworms also signify non-homologous phases, for example, the colour change of worms from black to swallow worms and diseases. Advances in artificial intelligence, especially deep learning techniques, and CNNs, have started using the techniques extensively and more effectively for image recognition of features. In contrast to the traditional approach using multivariate analysis as a feature extraction technique and classification method, deep learning algorithms use self-learning to recognize features across neural layers. By using various methods of discriminant analysis, such as deep learning, the study intends to create models of dead cocoons and create an automated system in the sericulture industry.