Add Over Saturation Effect on AVCaptureDevice in iOS

Question

I have implemented custom camera into my app. So, I am creating this camera using AVCaptureDevice. Also I have set GrayScale on this custom camera using Set GrayScale on Output of AVCaptureDevice in iOS this.

Now I want to add over saturation effect on camera. "Oversaturation hint" should be calculated according to formula provided. (convert overexposure pixel (R,G,B >= 255) to yellow (RGB 255,255,0)).

I want exact effect which is showing in following Image.

I am creating GrayScale using following code:-

let sepiaColor = CIColor(
    red: 1.0 / 0.30078125,
    green: 1.0 / 0.5859375,
    blue: 1.0 / 0.11328125
)

filter = CIFilter(
    name: "CIColorMonochrome",
    withInputParameters: [
        "inputColor" : sepiaColor,
        "inputIntensity" : 1.0
    ]
)

How can I create this over saturation effect using CIFilter.? I have reviewed Core Image Filter Reference document for CIFilter. But didn’t get idea about How to create this effect using this. May be CIColorCrossPolynomial or CIColorMap are working for this.

Is there any way to create this effect on AVCaptureDevice.?

Any Help would be appreciated. :)

I would create my own filter. You can follow these steps to do so: https://developer.apple.com/library/ios/documentation/GraphicsImaging/Conceptual/CoreImaging/ci_custom_filters/ci_custom_filters.html — Kunal Shrivastava, Jul 28 '16 at 13:24
This WWDC explains how to do that: https://developer.apple.com/videos/play/wwdc2014/514/ — Kunal Shrivastava, Jul 28 '16 at 14:00
Thanks @KunalShrivastava. Please provide any sample code for this. So it's good me. — Meet Doshi, Jul 28 '16 at 14:17
When you say "convert overexposure pixel (R,G,B >= 255) to yellow (RGB 255,255,0)", do you mean converting the pixels whose (R+G+B) valus is greater than or equal to than 255? — Kunal Shrivastava, Jul 28 '16 at 20:05

Kunal Shrivastava · Accepted Answer · 2016-07-28T20:41:17.810

Your logic/algorithm:

(convert overexposure pixel (R,G,B >= 255) to yellow (RGB 255,255,0))

will go in the kernel, which is essentially the computation you want to perform on each pixel.

Here's a simple filter I have written. In the if condition of the kernel, write your own condition. The then part already converts the pixel into (255, 255, 0).

.h

#import <CoreImage/CoreImage.h>

@interface CustomFilter : CIFilter

@property(nonatomic, retain) CIImage *inputImage;

-(CIImage*)outputImage;

@end

.m

#import "CustomFilter.h"

@implementation CustomFilter

-(CIColorKernel*)myKernel
{
    static CIColorKernel *kernel = nil;
    static dispatch_once_t once;
    dispatch_once(&once, ^{
        kernel = [CIColorKernel kernelWithString:
                  @"kernel vec4 CustomFilter ( __sample s ) \n { \n if ( s.r + s.g + s.b < 0.1 ) \n { return s.rgba = vec4(1.0, 1.0, 0.0, 1.0); } \n else \n { return s.rgba; } \n }"];
    });

    return kernel;
}

-(CIImage*)outputImage
{
    CGRect dod = _inputImage.extent;
    return [[self myKernel] applyWithExtent:dod arguments:@[_inputImage]];
}

@end

Great @kunal. It works. Thanks. I am using this condition s.r + s.g + s.b >= 1.0 instead of yours. Thank you again. — Meet Doshi, Jul 29 '16 at 06:56

Add Over Saturation Effect on AVCaptureDevice in iOS

1 Answers1