Vegetation region localization in natural environments is challenging due to uncontrolled factors such as overlapping, shadow, specular reflection, etc. A method is proposed using two feature-based clustering methods in complementary ways to localize the vegetation region in the field environment. Gaussian mixture model (GMM) and Fuzzy C-mean clustering (FCM) are both jointly used for the vegetation regions, where the first method provides weightage to feature similarity and the second method provides weightage to spatial constraints. Image region is segregated into three regions, further two regions combined based on higher intensity as well as similar features identified as vegetation regions. The proposed method is able to identify vegetation regions with F score 94.05%, MIoU 75.22%, Kappa 68.54% and accuracy 90.61% on AGM dataset, annotated for 200 images. Qualitative analysis is also performed on the maize field dataset.