- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
hi,
Is there any efficient way of performing arithmetic and other (warp actually) functions
on images (arrays) of different types?
for example - ippiSub_8u32f_..
ippiWarpPerspective_8u32f_..
etc.
I am looking for some implementation faster than convert before these functions.
Link Copied
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi,
Unfortunately, Intel IPP does not include such functions . You need to convert the image type first at this time.
So what is output data type, 8u, or 32f? We will forward it to Intel Image owners andsee if we can consider it in the future release.
Regards,
Chao
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi, sorry for the late response.
Most of my needs are for input of type 8u/16u (popular input images type) and output of type 32f, or 64 bit, used for further arithmetics and algorithms.
Hopefully will be inserted into future release..
thanx
lital
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
To be honest I do not think it will be significantly faster than conversion but it will add a lot of combination for data types. From our prospective, if you are going to implement some chain of high precision calculations it is better to convert image to appropriate data type.
Note, we keep increasing number of functions which support Ipp16u data type. Hope this help in your tasks.
Vladimir
- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page