Precision Agricultural Mapping: Enhancing Crop Edge Segmentation Through Satellite-Based Spatial DeepLabV3+

In crop edge segmentation, research on the different models has been done, which gives more accuracy than the previous approaches such as semantic segmentation, and MDSCBA segmentation. The merging of DeepLabV3+ and the spatial attention model is proposed for satellite-based crop edge segmentation. Modern linguistic segmentation methods like the DeepLabV3+ model can effectively separate the edge of crops from the rest of the landscape. The spatial attention model can selectively focus on informative regions and suppress irrelevant features, which can improve the accuracy of crop edge detection. This research suggests a new architecture in this study that incorporates the benefits of the two models. In comparison to the separate models, the suggested model performs better in terms of accuracy, precision, and recall. This study utilizes the DeepGlobe land cover dataset for this. The results of the research show that the suggested model is efficient at separating crop edges from satellite pictures and has potential applications in precision agriculture.