mxpi_semanticsegstitcher

Function	Merges the images of the semantic segmentation inference result.
Synchronous/Asynchronous (Status)	Synchronous
Constraints	It must be used together with the mxpi_roigenerator plugin.
Plugin Base Class (Factory)	mxpi_semanticsegstitcher
Input/Output	Input: three input ports. Port 1: Receives the inference result of the semantic segmentation model and metadata (data type: MxpiImageMaskList). Port 2: Receives the block object frame data generated by mxpi_roigenerator during block division; metadata (data type: MxpiObjectList). Port 3: Receives the target region for merging images generated by mxpi_roigenerator. (data type: MxpiObjectList) Output: metadata (data type: MxpiImageMaskList)
Port Format (Caps)	Static input: Port 1: {"metadata/semanticseg"} Port 2: {"metadata/object"} Port 3: {"metadata/object/roi-info"} Static output: {"metadata/object"}
Property	For details, see Table mxpi_semanticsegstitcher plugin properties.

**Table 1** mxpi_semanticsegstitcher plugin properties
Property Name	Description	Mandatory or Not	Modifiable or Not
dataSourceSemanticSeg	Index corresponding to the inference result of the semantic segmentation model. By default, it is automatically obtained based on the plugin connection status.	No	Yes
dataSourceCropRoi	Index of the block target frame data. By default, it is automatically obtained based on the plugin connection status.	No	Yes
dataSourceMergeRoi	Target area for image merging. By default, it is automatically obtained based on the plugin connection status.	No	Yes

Figure 1 Typical application process

Parent topic: Streaming Plugins