This study focuses on the development of a Weedy Rice Classification System using Convolutional Neural Networks (CNN). The study consists of several phases, starting with self-data collection involving images of weedy rice and cultivated rice from a few specific locations in Malaysia. Data augmentation techniques are applied to enhance the diversity of the dataset. The dataset is then split into training and testing sets for model development and evaluation. The CNN model achieves a promising accuracy of 91% during training. The trained model is implemented into the mobile application, and testing is conducted to evaluate the system's accuracy, revealing a performance of 85%. This system holds significant potential for automating the identification of weedy rice in agricultural settings, contributing to efficient crop management, and improving overall agricultural productivity. Finally, ongoing research and refinements are encouraged to further enhance the system's performance and attain even higher levels of accuracy.