Support Website Contact Support Blog

Image annotation and mask generation failure for non-RGB (e.g. Micasense) imagery

I have completed Intialization/Point Cloud/DSM processing steps for Micasense imagery (5 bands) and generated an ortho. However, almost all of the images contain our drone’s shadow, resulting in a speckling of drone shadows across the stitched mosaic. I want to set a mask to remove the drone in each of the images. In following the steps outlined at https://support.pix4d.com/hc/en-us/articles/202560549-How-to-Annotate-Images-in-the-rayCloud I receive a Pix4D error when selecting the annotation tool. Is it not possible to mask/annotate Micasense imagery? Is this a bug?

I am new to Pix4D but suspect the annotation tool works well for RGB imagery and can use the results of RGB annotation to craft a solution for Micasense data. I am wondering if I can use the files generated through the mask process (https://community.pix4d.com/t/8276-Image-Annotation-Removal) from a set of RGB images. I would then use these files as a basis to learn how to create segmentation/mask files by hand (for Micasense images). Anyone try this?

Thank you!

Hi Drew,

Indeed, multispectral images cant be masked or annotated. There is no such tool to remove some pixels in our software. However, after processing you can use any other tools/languages like Matlab/ Python/ C++ and replace these values with nearby pixel values that are pure and are not having those effects.