Dear all, I have done some experiments to verify if the parameters of for the lens of Canon EF 75-300mm f/4-5.6 III are good.
I used two programs: one is the in ~/lensfun-0.2.7/tests/tmod, and the other one is that I wrote by myself. Two images found on the internet are tried, both are taken by the lens Canon EF 75-300mm f/4-5.6 III. One is taken with the focal length 90mm, and the other one is taken with 100mm. (PS: I would like also to find one image taken with the focal length 95mm. But I didn't find one.) The corresponding distortion model and parameters are used to correct the distortion. The two programs I tried give the very similar results. You can see the images in the following links: original image 1: https://dl.dropboxusercontent.com/u/68869086/Canon_EF_75-300mm_f4-5.6III_focal90mm.jpg corrected image 1: https://dl.dropboxusercontent.com/u/68869086/out-Canon_EF_75-300mm_f4-5.6III_focal90mm.png original image 2: https://dl.dropboxusercontent.com/u/68869086/Canon_EF_75-300mm_f4-5.6III_focal100mm_roof.jpg corrected image 2: https://dl.dropboxusercontent.com/u/68869086/out-Canon_EF_75-300mm_f4-5.6III_focal100mm_roof.png As "mbg" suggests, since the curve I show at the link above is almost linear in the portion when Rd < Ru, we can also stay in the image domain where Ru < 1. But as we can see from the images, even the domain close to the center of image is still not corrected. So I think the distortion parameters of this Canon lens is problematic. Anyone has some comments ? thank you ------------------------------------------------------------------------------ Put Bad Developers to Shame Dominate Development with Jenkins Continuous Integration Continuously Automate Build, Test & Deployment Start a new project now. Try Jenkins in the cloud. http://p.sf.net/sfu/13600_Cloudbees _______________________________________________ Lensfun-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/lensfun-users
