In this research, diseases and micronutrient deficiencies were shown to influence sugarcane productivity greatly and, therefore, should be promptly enhanced productivity. Orchestration techniques like personnel check-ups and lab analyses are repetitive, require a lot of time, and may involve human error-to-date interventions. Emerging technologies in precision agriculture have applied DL to address these problems. From Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs) to hybrid architectures ResNet and VGGNet, significant progress has been made in boosting disease diagnosis and micronutrient deficiencies diagnosis in sugarcane crops. However, in these models, the dataset size and the computational requirement of real-time applications persist. This review discusses DL strategies on image data based on Conventional CNN, temporal data based on RNN, and improving DL’s robustness by combining CNN and RNN architectures. It also reviews picture pre-processing, picture segmentation and feature extraction for higher overall picture classification performance. This review of 50+ research articles from 2015 to 2024 examines the existing knowledge gap, the prospects for future AI developments, and AI’s potential in sugarcane farming. The study presents issues such as limited datasets and real-time functionality to provide an outline for subsequent research on AI-facilitated precision agriculture.