<p>Agriculture supports human survival and contributes to the Indian economy. To improve agricultural productivity, selecting a suitable agricultural crop based on the soil requirements is significant. So, a smart agricultural monitoring system based on IoT networking is proposed for a crop recommendation system in this research to increase the crop yield of the land, which would benefit the agriculturalists. The existing crop recommendation systems suffer from lower recommendation accuracy, computational complexities, issues in data quality, delay, inaccuracies in the data collection, and so on. In this research, an IoT-based smart agricultural system is implemented using the proposed Hybrid Activation-enabled Neural Network-based Adaptive Velocity Spiral Optimization (HANN-Ada-VeSO) model to recommend the crops based on the land regions and soil properties. The efficiency of the proposed HANN-Ada-VeSO model is accomplished by integrating hybrid activation functions and the Adaptive Velocity Spiral Optimization (Ada-VeSO) algorithm in the Two-Layered Neural Network model, which empowers to recommend the suitable crops accurately. Additionally, the Ada-VeSO algorithm tunes the parameters of the Hybrid Activation-enabled Neural Network (HANN) model, which improves the performance in recommending the crops and solves various issues like computational complexities. The outcomes obtained by the HANN-Ada-VeSO model show that it reaches an accuracy of 94.61%, precision of 93.84%, recall of 95.14%, F1-Score of 96.41%, specificity of 93.84%, and NPV of 95.14%, respectively.</p>