Semantic Segmentation under a Complex Background for Machine Vision Detection Based on Modified UPerNet with Component Analysis Modules

Figure 2

Flowchart of the semantic segmentation under a complex background using the encoder-decoder network.