Recognizing Zucchinis Intercropped with Sunflowers in UAV Visible Images Using an Improved Method Based on OCRNet

An improved semantic segmentation method based on object contextual representations network (OCRNet) is proposed to accurately identify zucchinis intercropped with sunflowers from unmanned aerial vehicle (UAV) visible images taken over Hetao Irrigation District, Inner Mongolia, China. The improves the performance of OCRNet in two respects. First, region context extraction structure OCRNet, a branch that uses channel attention module was added parallel rationally use feature maps different weights and reduce noise invalid features. Secondly, Lovász-Softmax loss introduced improve accuracy representation optimize final result at level. We compared extant advanced methods (PSPNet, DeepLabV3+, DNLNet, OCRNet) test areas its effectiveness. results showed achieved best effect areas. More specifically, our performed better processing image details, segmenting field edges, identifying intercropping fields. has significant advantages for crop classification recognition UAV images, these are more substantive object-level evaluation metrics (mIoU IoU).