One of the most damaging agricultural pests that seriously affect crops is weeds. Weeds cause an increase in the cost of cultivation owing to the wastage of crops which causes a large economic loss in agriculture worldwide. Because of the importance of this problem, experts are looking into how technology may help farmers spot weeds early. This paper focuses on various weeds which are identical to that of crops. They can be detected by making use of crop field photos and classifying them with artificial intelligence-driven visualization. The work analyzed images taken from drones, IoT sensors, and phones using machine learning and deep learning techniques. For analysis of the work, ResNet50 model results have been evaluated and compared with different algorithms based on three evaluation parameters namely accuracy, recall, and F1-Score. The results show that ResNet50 works best in terms of all parameters, whereas VGG19 becomes the second-best-performing model, and VGG16 performs moderately well but is less efficient than the other two models. At last, DenseNet121 has the highest precision but the lowest F1-Score due to its lower accuracy and recall percentage.