When trying to fetch dominant colors from some images an exception is raised.

File: DominantColors.swift line 159

This is the image that's causing crash

In my code I just called the dominantColors() method with default arguments.

I think this might be a bug? Here's my code.

let greenComp:UIColor = UIColor.green.complementaryColor //works let greenCompComp:UIColor = greenComp.complimentaryColor //error: value of type 'UIColor' has no member 'complimentaryColor'

I know that UIColor does have member complimentaryColor cause it just used it in line 1. Not sure if I did something wrong.

With certain images the primary/backgroundColor values could not be obtained which led to crash when trying to unwrap those values.

In such cases debugger prints such message: -[CIContext render:toBitmap:rowBytes:bounds:format:colorSpace:] unsupported colorspace.

Hey, first of all thanks for this amazing, impressive library.

I'd like to report that when using .kMeansClustering algorithm, I often get a crash from unwrapping nil value. It happens at line 112 in file ColorPalette.swift

The image I'm feeding to the algorithm is a transparent PNG (an asset file) so that's probably the cause of it. I know it doesn't make sense to pull colors from such image but the code should fail gracefully in such cases.

To fix the warning we get when running workflows: Node.js 12 actions are deprecated. For more information see: https://github.blog/changelog/2022-09-22-github-actions-all-actions-will-begin-running-on-node16-instead-of-node12/. Please update the following actions to use Node.js 16

The name() function introduced recently was only supporting English. It is now setup to use Apple's localization system. This PR also adds the first localized language: French.

Fixes #11.

Via #10, we introduced a new function called name() that gives the English name of a UIColor instance.

Localizing those names makes sense as a next step.

This commit simply adds a new function name() to UIColor instances via an extension. This function is useful to get a readable english name of the current color.

Fixes #6.

Expected

let colors = try image.dominantColors()
let palette = ColorPalette(orderedColors: colors, ignoreContrastRatio: true)

Actual

let colors = try image.dominantColors()
let palette = ColorPalette(orderedColors: colors ignoreContrastRatio: true)

A useful feature for accessibility is to be able to attach a readable color name to UIColor instances. For example UIColor.red or UIColor(red: 1.0, green: 0.0, blue: 0.0) should read something like "red".

A trivial implementation of this could be to work with a predefined color palette and just compare it to the colors using the existing deltaE.

I'm using this function on each image of a video feed, and after a few seconds/tens of second, my program stops with this error, EXC_BAD_ACCESS (code=1, address=0xd833deece4d0). I assume it comes from this function and precisely kMeansClustering, because I don't encounter this issue using .iterative algorithm

Hello,

first of all I wanted to say thank you for this awesome kit.

I'm using ColorKit to generate a color palette of pictures that I manually add from my photo library. Everything work good but I only have problems with pictures of my library that were not taken with my iPhone.

If I want to get the colors of pictures taken with my iPhone or a screenshot which was taken with my iPhone the function .dominantColors does not work.

Is it a format thing? or maybe I missed something.

If I send the picture per WhatsApp and then I download the same picture I can get the colors but not from the original. Photos taken with an external camera and then imported to the iPhone have also no problem at all. Only Iphone photos.

I appreciate your help, thank you very much!

Pablo.

UPDATE I think that the problem is that ColorKit can't use .dominantColors() with pictures taken with the iPhone because they are HEIF/HEIC files. I still can't solve the problem but I'm trying to convert the image to JPEG. I don't know if that is useful because I thought UIImage also supports HEIF format.

Generating color histograms out of images.

Here is some good resource from OpenCV as to what type of histogram we could generate as well as how to represent them.