Over the years, weed growth has evolved into a significant component that influences overall agricultural productivity and mostly incur to losses. Hence, timely weed detection and control can have a significant favorable impact on the productivity of crops and can provide valuable insights for precision farming. By using the concepts and techniques of machine learning, weeds can be easily identified and categorized. Convolutional Neural Network, one of the deep learning techniques, when experimented with led to better results than the majority of the others. In this work, we develop a CNN model that can be applied to weed pictures to recognize as well as categorize them depending upon their type. By using this model, one could considerably raise the effectiveness and efficiency of weed control in soybean crops.