Optimized deep learning multi-model investigation of images for ground water level detection

Abstract

<p>When referring to groundwater in the context of the normal water cycle, the word “reservoir” is acceptable. Compared to the atmosphere or fresh surface water, groundwater has a much greater capacity to hold water. Despite the high degree of variety and complexity of the subsurface ecosystem, there is currently only a small amount of data available. People who depended on reality-based models ran into both of these challenges at some point. Statistical modeling was used in order to get a more accurate calibration of the model throughout the course of time. Because of the expansion in global population, governments in both wealthy and developing nations are increasingly looking to groundwater as an essential resource for satisfying the water requirements of their populations. Because water was preserved in such a large amount, it may be used again, even when there is a drought or other dry time. In this manuscript, a Particle Swarm Optimization (PSO) enabled Visual Geometry Group (VGG) 16 deep learning model is used in order to determine the level of ground water. Accuracy, specificity and sensitivity of PSO enabled VGG is highest among the algorithms used in the experimental study. Accuracy of PSO VGG 16 is 92.5 percent, specificity and sensitivity of PSO VGG 16 is 99 percent.</p>