Compiling pHash for iOS

Question

How to compile pHash for iOS?

I read the documentation but there's no mention of iOS/arm and Google doesn't help.

So if anyone was able to compile pHash for iOS please share your experience.

Bechler

@CodaFi I managed to compile for Mac with Homebrew and applying a patch from Wulfshade – Bechler

Can be somewhat tedious, see my post here: #11453212 – Colorful

@AnriDigholm +1 for that, but I managed to compile for Mac with Homebrew - the problem is iOS – Bechler

pHash and any GPL-ed component cannot be used of iOS because the App Store terms are incompatible with GPL. So, you better put your efforts into finding an alternative... – Fluxmeter

GPL is not a problem - pHASH also has a commercial license available. – Bechler

@AndreiStoleru did you succeeded ? – Fibered

@Fibered actually no. In the end I "re-wrote" pHash in Objective-C (with support for Accelerate framework) – Bechler

@AndreiStoleru did you open sourced it ? :D – Fibered

@Fibered no, it was for a client. I'll ask my client if he's ok with open sourcing the code – Bechler

@AndreiStoleru ok cool :) I think that's not hard to convert it. I just need the audio part so I can ignore CImg things. – Fibered

@Fibered My library will not be of much use in this case. I wrote it with perceptual hashing of images in mind and all the calculations are based on this :( – Bechler

@AndreiStoleru of no problem :) I'll write it for audio hash myself. Thx ! – Fibered

Answer 1

If you need pHash for imageHash purposes, You might consider looking into the imageMagick framework for iOS : http://www.imagemagick.org/download/iOS/

Also, you should remember that pHash is licensed under GPLv3, which means that derived work can only be distributed under the same license terms.

Rosser

pHash also has a commercial license so GPL is not a problem. I had a look at imagemagick but it seems it has the same problems when it comes to iOS. – Bechler

imageMagick does image hash for exact identical pixels, while pHash is a project that could compare the hash of resized, distorsion, cropping images. – Ossicle