Hello,
Thank you for this great tool! I have a few feature requests and questions regarding the QGIS plugin:
1. Feature Request: Enhance output Shapefile attributes & default save path
Currently, when using point, box, or text prompts and setting the output format to .shp, the output lacks some contextual information.
Auto-add attributes: I would like to request adding Name (the filename of the source image being segmented) and Type (the type of prompt used, e.g., point, box, text) attributes to the output .shp file.
Auto-save path: Automatically default the save path of the generated .shp file to the same local directory as the source image.
Note: Is it possible to implement this in a future release? Alternatively, if you could point me to the relevant sections of the codebase, I would appreciate any guidance on how to modify the code to achieve this myself.
2. Feature Request: SAM 1/SAM 2 specific weights
It would be great to have the ability to select specific model weights. Such as ViT-B, Vit-L for SAM 1 and small, large for SAM 2.
3. Question: Best practices for large remote sensing images (.tif)
What is the recommended workflow for handling large .tif files (e.g., a full remote sensing scene)? When inputting an entire large image, the SAM segmentation results are usually poor. My current workaround is to crop the large image into smaller tiles, segment them individually, and then merge the results.
Are there better processing methods or any built-in tile-and-merge features planned for this plugin?
Thank you very much for your time and assistance!
Hello,
Thank you for this great tool! I have a few feature requests and questions regarding the QGIS plugin:
1. Feature Request: Enhance output Shapefile attributes & default save path
Currently, when using point, box, or text prompts and setting the output format to .shp, the output lacks some contextual information.
Auto-add attributes: I would like to request adding Name (the filename of the source image being segmented) and Type (the type of prompt used, e.g., point, box, text) attributes to the output .shp file.
Auto-save path: Automatically default the save path of the generated .shp file to the same local directory as the source image.
Note: Is it possible to implement this in a future release? Alternatively, if you could point me to the relevant sections of the codebase, I would appreciate any guidance on how to modify the code to achieve this myself.
2. Feature Request: SAM 1/SAM 2 specific weights
It would be great to have the ability to select specific model weights. Such as ViT-B, Vit-L for SAM 1 and small, large for SAM 2.
3. Question: Best practices for large remote sensing images (.tif)
What is the recommended workflow for handling large .tif files (e.g., a full remote sensing scene)? When inputting an entire large image, the SAM segmentation results are usually poor. My current workaround is to crop the large image into smaller tiles, segment them individually, and then merge the results.
Are there better processing methods or any built-in tile-and-merge features planned for this plugin?
Thank you very much for your time and assistance!