In several research papers several authors have used to convert RAW file format to JPEG and process with Pix4Dmapper to produce the cloud point. The JPEG is a compress file format with several degree of compression (loosing image information). Why is it so common to use JPEG instead of TIFF format, where there is not compression?
I have tried both ways and there is a slight improvement by going to tiff, but not worth the overhead of the file size, conversion time and the time process the larger files. I now shoot RAW and JPEG, but only use the RAW if there is a conversion issue. I have not tried TIFF for close in modeling however. Most of my work is aerial mapping.
I have been tempted to convert to TIFF as well only yesterday after having too much noise on DSMs I produced from JPEGs. Now, if you are saying there is only little improvement, that kind of puts me down. I’m at this stage very frustrated with bad DSM I get from surveying waste dump, stockpiles and the pit around the mine. There is a bit vegetation around the dump and stockpiles mostly grass, outside these areas there is dense forest, which I should not worry about, but what I’m yet to understand the poor quality of DSM I get around mostly open areas. I’m now also tempted to turn of filter and smoothing. In the mean time I will continue testing TIFF, each image is about 100Mb, conversion time from RAW was not significantly different compared to conversion to JPEG, but JPEG file size was about 245Mb
Thanks for this article.
I have moved to using RAW images (converted to TIF) for creating agricultural NDVI maps. I do find improvements in the index maps obtained - that is, more sensitivity to changes in crop vigour.