Related
I am trying to create an image mask with kCGColorSpaceDisplayP3 colorspace to support the iPhone 7's wide color range.
I am able to create image mask correctly when using sRGB colorspace on iPhone 6 and earlier devices using iOS 10 and earlier iOS. But I have no clue where I am going wrong when creating colorspace using kCGColorSpaceDisplayP3:
CGColorSpaceRef colorSpace = CGColorSpaceCreateWithName(kCGColorSpaceDisplayP3);
CGContextRef context = CGBitmapContextCreate(NULL, 320.0, 320.0, 32, 320.0*16, colorSpace, kCGImageAlphaPremultipliedLast | kCGBitmapFloatComponents);
CGFloat radius = 10.0;
CGFloat components[] = {1.0,1.0,1.0,1.0, 1.0,1.0,1.0,1.0, 1.0,1.0,1.0,1.0, 1.0,1.0,1.0,1.0, 1.0,1.0,1.0,0.5, 1.0,1.0,1.0,0.0};
CGFloat locations[] = {0.0, 0.1, 0.2, 0.8, 0.9, 1.0};
CGGradientRef gradient = CGGradientCreateWithColorComponents(colorSpace, components, locations, 6); //colorSpaceP3
CGPoint center = CGPointMake(100.0, 100.0);
CGContextDrawRadialGradient(context, gradient, center, 0.1, center, radius, 0);
CGGradientRelease(gradient);
CGImageRef imageHole = CGBitmapContextCreateImage(context);
CGImageRef maskHole = CGImageMaskCreate(CGImageGetWidth(imageHole), CGImageGetHeight(imageHole), CGImageGetBitsPerComponent(imageHole), CGImageGetBitsPerPixel(imageHole), CGImageGetBytesPerRow(imageHole), CGImageGetDataProvider(imageHole), NULL, FALSE);
CGImageRelease(imageHole);
CGImageRef image = [UIImage imageNamed:#"prosbo_hires.jpg"].CGImage;
CGImageRef masked = CGImageCreateWithMask(image, maskHole);
CGImageRelease(maskHole);
UIImage *img = [UIImage imageWithCGImage:masked];
CGImageRelease(masked);
CGContextRelease(context);
CGColorSpaceRelease(colorSpace);
The log says:
: CGImageMaskCreate: invalid mask bits/component: 32.
I don't have much experience with Core Graphics. Can anyone please suggest something here.
Thanks.
The documentation for the bitsPerComponent parameter of CGImageMaskCreate() says:
Image masks must be 1, 2, 4, or 8 bits per component.
You're passing CGImageGetBitsPerComponent(imageHole), which is 32 bits per component. As per both the documentation and the log message, that's not valid.
The implication is that image masks don't support floating point bitmap formats.
It should be possible to create the bitmap context and the mask using 8 bits per component. More or less, just leave out kCGBitmapFloatComponents. I expect that will result in reduced granularity of the opacity of the mask, but won't affect the color range of masked images.
this fixed my issue:
contextRef = CGBitmapContextCreate(
m.data,
m.cols,
m.rows,
8,
m.step[0],
CGColorSpaceCreateDeviceRGB(),
bitmapInfo);
https://developer.apple.com/search/?q=CGColorSpaceCreate
Is there any way to draw an NSImage like images in NSButtons or other cocoa interface elements?
Here are examples:
Apple uses pdf's with black icons:
If you simply want this effect to be applied when you use your own images in a button, use [myImage setTemplate:YES]. There is no built-in way to draw images with this effect outside of a button that has the style shown in your screenshots.
You can however replicate the effect using Core Graphics. If you look closely, the effect consists of a horizontal gradient, a white drop shadow and a dark inner shadow (the latter is the most difficult).
You could implement this as a category on NSImage:
//NSImage+EtchedDrawing.h:
#interface NSImage (EtchedImageDrawing)
- (void)drawEtchedInRect:(NSRect)rect;
#end
//NSImage+EtchedDrawing.m:
#implementation NSImage (EtchedImageDrawing)
- (void)drawEtchedInRect:(NSRect)rect
{
NSSize size = rect.size;
CGFloat dropShadowOffsetY = size.width <= 64.0 ? -1.0 : -2.0;
CGFloat innerShadowBlurRadius = size.width <= 32.0 ? 1.0 : 4.0;
CGContextRef c = [[NSGraphicsContext currentContext] graphicsPort];
//save the current graphics state
CGContextSaveGState(c);
//Create mask image:
NSRect maskRect = rect;
CGImageRef maskImage = [self CGImageForProposedRect:&maskRect context:[NSGraphicsContext currentContext] hints:nil];
//Draw image and white drop shadow:
CGContextSetShadowWithColor(c, CGSizeMake(0, dropShadowOffsetY), 0, CGColorGetConstantColor(kCGColorWhite));
[self drawInRect:maskRect fromRect:NSMakeRect(0, 0, self.size.width, self.size.height) operation:NSCompositeSourceOver fraction:1.0];
//Clip drawing to mask:
CGContextClipToMask(c, NSRectToCGRect(maskRect), maskImage);
//Draw gradient:
NSGradient *gradient = [[[NSGradient alloc] initWithStartingColor:[NSColor colorWithDeviceWhite:0.5 alpha:1.0]
endingColor:[NSColor colorWithDeviceWhite:0.25 alpha:1.0]] autorelease];
[gradient drawInRect:maskRect angle:90.0];
CGContextSetShadowWithColor(c, CGSizeMake(0, -1), innerShadowBlurRadius, CGColorGetConstantColor(kCGColorBlack));
//Draw inner shadow with inverted mask:
CGColorSpaceRef colorSpace = CGColorSpaceCreateDeviceRGB();
CGContextRef maskContext = CGBitmapContextCreate(NULL, CGImageGetWidth(maskImage), CGImageGetHeight(maskImage), 8, CGImageGetWidth(maskImage) * 4, colorSpace, kCGImageAlphaPremultipliedLast);
CGColorSpaceRelease(colorSpace);
CGContextSetBlendMode(maskContext, kCGBlendModeXOR);
CGContextDrawImage(maskContext, maskRect, maskImage);
CGContextSetRGBFillColor(maskContext, 1.0, 1.0, 1.0, 1.0);
CGContextFillRect(maskContext, maskRect);
CGImageRef invertedMaskImage = CGBitmapContextCreateImage(maskContext);
CGContextDrawImage(c, maskRect, invertedMaskImage);
CGImageRelease(invertedMaskImage);
CGContextRelease(maskContext);
//restore the graphics state
CGContextRestoreGState(c);
}
#end
Example usage in a view:
- (void)drawRect:(NSRect)dirtyRect
{
[[NSColor colorWithDeviceWhite:0.8 alpha:1.0] set];
NSRectFill(self.bounds);
NSImage *image = [NSImage imageNamed:#"MyIcon.pdf"];
[image drawEtchedInRect:self.bounds];
}
This would give you the following result (shown in different sizes):
You may need to experiment a bit with the gradient colors and offset/blur radius of the two shadows to get closer to the original effect.
If you don't mind calling a private API, you can let the operating system (CoreUI) do the shading for you. You need a few declarations:
typedef CFTypeRef CUIRendererRef;
extern void CUIDraw(CUIRendererRef renderer, CGRect frame, CGContextRef context, CFDictionaryRef object, CFDictionaryRef *result);
#interface NSWindow(CoreUIRendererPrivate)
+ (CUIRendererRef)coreUIRenderer;
#end
And for the actual drawing:
CGRect drawRect = CGRectMake(x, y, width, height);
CGImageRef cgimage = your_image;
CFDictionaryRef dict = (CFDictionaryRef) [NSDictionary dictionaryWithObjectsAndKeys:
#"backgroundTypeRaised", #"backgroundTypeKey",
[NSNumber numberWithBool:YES], #"imageIsGrayscaleKey",
cgimage, #"imageReferenceKey",
#"normal", #"state",
#"image", #"widget",
[NSNumber numberWithBool:YES], #"is.flipped",
nil];
CUIDraw ([NSWindow coreUIRenderer], drawRect, cg, dict, nil);
CGImageRelease (cgimage);
This will take the alpha channel of cgimage and apply the embossing effect as seen on toolbar buttons. You may or may not need the "is.flipped" line. Remove it if your result is upside-down.
There are a bunch of variations:
kCUIPresentationStateKey = kCUIPresentationStateInactive: The window is not active, the image will be lighter.
state = rollover: Only makes sense with the previous option. This means you are hovering over the image, the window is inactive, but the button is sensitive (click-through is enabled). It will become darker.
state = pressed: Occurs when the button is pressed. The icon gets slightly darker.
Bonus tip: To find out stuff like this, you can use the SIMBL plugin CUITrace. It prints out all the CoreUI invocations of a target app. This is a treasure trove if you have to draw your own native-looking UI.
Here's a much simpler solution: just create a cell and let it draw. No mucking around with private APIs or Core Graphics.
Code could look similar to the following:
NSButtonCell *buttonCell = [[NSButtonCell alloc] initImageCell:image];
buttonCell.bordered = YES;
buttonCell.bezelStyle = NSTexturedRoundedBezelStyle;
// additional configuration
[buttonCell drawInteriorWithFrame: someRect inView:self];
You can use different cells and configurations depending on the look you want to have (eg. NSImageCell with NSBackgroundStyleDark if you want the inverted look in a selected table view row)
And as a bonus, it will automatically look correct on all versions of OS X.
To get to draw correctly within any rect, the CGContextDrawImage and CGContextFillRect for the inner mask must have the origin of (0,0). then when you draw the image for the inner shadow you can then reuse the mask rect. So ends up looking like:
CGRect cgRect = CGRectMake( 0, 0, maskRect.size.width, maskRect.size.height );
CGColorSpaceRef colorSpace = CGColorSpaceCreateDeviceRGB();
CGContextRef maskContext = CGBitmapContextCreate( NULL, CGImageGetWidth( maskImage ), CGImageGetHeight( maskImage ), 8, CGImageGetWidth( maskImage ) * 4, colorSpace, kCGImageAlphaPremultipliedLast );
CGColorSpaceRelease( colorSpace );
CGContextSetBlendMode( maskContext , kCGBlendModeXOR );
CGContextDrawImage( maskContext, cgRect, maskImage );
CGContextSetRGBFillColor( maskContext, 1.0, 1.0, 1.0, 1.0 );
CGContextFillRect( maskContext, cgRect );
CGImageRef invertedMaskImage = CGBitmapContextCreateImage( maskContext );
CGContextDrawImage( context, maskRect, invertedMaskImage );
CGImageRelease( invertedMaskImage );
CGContextRelease( maskContext );
CGContextRestoreGState( context );
You also have to leave a 1px border around the outside of the image or the shadows won't work correctly.
I have a UIImage I draw lines on it that follow the users finger. It's like a drawing board. This works perfectly when the UIImage is small, say 500 x 600, but if it's like 1600 x 1200, it gets really scratchy and laggy. Is there a way I can optimize this? This is my drawing code in touchesModed:
UIGraphicsBeginImageContext(self.frame.size);
[drawImageView.image drawInRect:CGRectMake(0, 0, self.frame.size.width, self.frame.size.height)];
CGContextSetLineCap(UIGraphicsGetCurrentContext(), kCGLineCapRound);
CGContextSetLineWidth(UIGraphicsGetCurrentContext(), 15.0);
CGContextSetRGBStrokeColor(UIGraphicsGetCurrentContext(), 1.0, 0.0, 0.0, 1.0);
//CGContextSetAlpha(UIGraphicsGetCurrentContext(), 0.5f);
CGContextBeginPath(UIGraphicsGetCurrentContext());
CGContextMoveToPoint(UIGraphicsGetCurrentContext(), lastPoint.x, lastPoint.y);
CGContextAddLineToPoint(UIGraphicsGetCurrentContext(), currentPoint.x, currentPoint.y);
CGContextStrokePath(UIGraphicsGetCurrentContext());
drawImageView.image = UIGraphicsGetImageFromCurrentImageContext();
UIGraphicsEndImageContext();
Thanks.
Instead of drawing the entire 1600X1200 frame at a single go, why not draw it as it's needed. What I mean is since you are drawing the entire frame (which resides in memory) you are having patchy performance.
Try CATiledLayer. Basically you need to draw a screen size only which your device is capable of, anything more than that as the user scrolls you need to draw it on the fly.
This is what is used in Google Maps on iPad or iPhone. Hope this helps...
Instead of creating a new context and drawing the current image into it every time a touch moves, create a context using CGBitmapContextCreate and reuse that context. All previous drawing will already be in the context, and you won't have to create a new context each time a touch moves.
- (CGContextRef)drawingContext {
if(!context) { // context is an instance variable of type CGContextRef
CGColorSpaceRef colorSpace = CGColorSpaceCreateDeviceRGB();
if(!colorSpace) return nil;
context = CGBitmapContextCreate(NULL, contextSize.width, contextSize.height,
8, contextSize.width * 32, colorSpace,
kCGImageAlphaPremultipliedFirst);
CGColorSpaceRelease(colorSpace);
if(!context) return nil;
CGContextConcatCTM(context, CGAffineTransformMake(1,0,0,-1,0,contextSize.height));
CGContextDrawImage(context, (CGRect){CGPointZero,contextSize}, drawImageView.image.CGImage);
CGContextSetLineCap(UIGraphicsGetCurrentContext(), kCGLineCapRound);
CGContextSetLineWidth(UIGraphicsGetCurrentContext(), 15.0);
CGContextSetRGBStrokeColor(UIGraphicsGetCurrentContext(), 1.0, 0.0, 0.0, 1.0);
}
return context;
}
- (void)touchesMoved:(NSSet *)touches withEvent:(UIEvent *)event {
CGContextRef ctxt = [self drawingContext];
CGContextBeginPath(ctxt);
CGContextMoveToPoint(ctxt, lastPoint.x, lastPoint.y);
CGContextAddLineToPoint(ctxt, currentPoint.x, currentPoint.y);
CGContextStrokePath(ctxt);
CGImageRef img = CGBitmapContextCreateImage(ctxt);
drawImageView.image = [UIImage imageWithCGImage:img];
CGImageRelease(img);
}
This code requires the instance variables CGContextRef context and CGSize contextSize. The context will need to be released in dealloc, and whenever you change its size.
I am trying to draw to a bitmap context but coming up empty. I believe I'm creating things properly because I can initialize the context, draw a few things, then create an image from it and draw that image. What I cannot do is, after initialization, trigger further drawing on the context that draws more items on it. I'm not sure if I'm missing some common practice that implies I can only draw it at certain places or that I have to do something else. Here is what I do, below.
I copied the helper function provided by apple with one modification to obtain the color space because it wasn't compiling (this is for iPad, don't know if that matters):
CGContextRef MyCreateBitmapContext (int pixelsWide, int pixelsHigh)
{
CGContextRef context = NULL;
CGColorSpaceRef colorSpace;
void * bitmapData;
int bitmapByteCount;
int bitmapBytesPerRow;
bitmapBytesPerRow = (pixelsWide * 4);// 1
bitmapByteCount = (bitmapBytesPerRow * pixelsHigh);
colorSpace = CGColorSpaceCreateDeviceRGB(); //CGColorSpaceCreateWithName(kCGColorSpaceGenericRGB);// 2
bitmapData = malloc( bitmapByteCount );// 3
if (bitmapData == NULL)
{
fprintf (stderr, "Memory not allocated!");
return NULL;
}
context = CGBitmapContextCreate (bitmapData,// 4
pixelsWide,
pixelsHigh,
8, // bits per component
bitmapBytesPerRow,
colorSpace,
kCGImageAlphaPremultipliedLast);
if (context== NULL)
{
free (bitmapData);// 5
fprintf (stderr, "Context not created!");
return NULL;
}
CGColorSpaceRelease( colorSpace );// 6
return context;// 7
}
I initialize it in my init method below with a few sample draws just to be sure it looks right:
mContext = MyCreateBitmapContext (rect.size.width, rect.size.height);
// sample fills
CGContextSetRGBFillColor (mContext, 1, 0, 0, 1);
CGContextFillRect (mContext, CGRectMake (0, 0, 200, 100 ));
CGContextSetRGBFillColor (mContext, 0, 0, 1, .5);
CGContextFillRect (mContext, CGRectMake (0, 0, 100, 200 ));
CGContextSetRGBStrokeColor(mContext, 1.0, 1.0, 1.0, 1.0);
CGContextSetRGBFillColor(mContext, 0.0, 0.0, 1.0, 1.0);
CGContextSetLineWidth(mContext, 5.0);
CGContextAddEllipseInRect(mContext, CGRectMake(0, 0, 60.0, 60.0));
CGContextStrokePath(mContext);
In my drawRect method, I create an image from it to render it. Maybe I should be creating and keeping this image as a member var and updating it everytime I draw something new and not creating the image every frame? (Some advice on this would be nice):
// draw bitmap context
CGImageRef myImage = CGBitmapContextCreateImage (mContext);
CGContextDrawImage(context, rect, myImage);
CGImageRelease(myImage);
Then as a test I try drawing a circle when I touch, but nothing happens, and the touch is definitely triggering:
- (void)touchesBegan:(NSSet *)touches withEvent:(UIEvent *)event
{
CGPoint location;
for (UITouch* touch in touches)
{
location = [touch locationInView: [touch view]];
}
CGContextSetRGBStrokeColor(mContext, 1.0, 1.0, 1.0, 1.0);
CGContextSetRGBFillColor(mContext, 0.0, 0.0, 1.0, 1.0);
CGContextSetLineWidth(mContext, 2.0);
CGContextAddEllipseInRect(mContext, CGRectMake(location.x, location.y, 60.0, 60.0));
CGContextStrokePath(mContext);
}
Help?
[self setNeedsDisplay]; !!!!
!!!!!!!!!
so it was because drawRect was never being called after init since it didn't know it needed to refresh. My understanding is that I should just call setNeedsDisplay anytime I draw and that seems to work. :)
So I tried to use the Quartz CGImageCreateWithMaskingColors function, but he only problem is that it masks the color range you are selecting.
I want to mask everything but the color range I am selecting. For instance, I want to show all red colors in a picture but remove the other channels (Green and Blue).
I am doing this in Objective-C and I am a noob so please give me examples :)
Any help is greatly appreciated.
use these methods.i found them in one of the SO posts.
-(void)changeColor
{
UIImage *temp23=Image;//Pass your image here
CGImageRef ref1=[self createMask:temp23];
const float colorMasking[6] = {1.0, 3.0, 1.0, 2.0, 2.0, 3.0};
CGImageRef New=CGImageCreateWithMaskingColors(ref1, colorMasking);
UIImage *resultedimage=[UIImage imageWithCGImage:New];
EditImageView.image = resultedimage;
[EditImageView setNeedsDisplay];
}
-(CGImageRef)createMask:(UIImage*)temp
{
CGImageRef ref=temp.CGImage;
int mWidth=CGImageGetWidth(ref);
int mHeight=CGImageGetHeight(ref);
int count=mWidth*mHeight*4;
void *bufferdata=malloc(count);
CGColorSpaceRef colorSpaceRef = CGColorSpaceCreateDeviceRGB();
CGBitmapInfo bitmapInfo = kCGBitmapByteOrderDefault;
CGColorRenderingIntent renderingIntent = kCGRenderingIntentDefault;
CGContextRef cgctx = CGBitmapContextCreate (bufferdata,mWidth,mHeight, 8,mWidth*4, colorSpaceRef, kCGImageAlphaPremultipliedFirst);
CGRect rect = {0,0,mWidth,mHeight};
CGContextDrawImage(cgctx, rect, ref);
bufferdata = CGBitmapContextGetData (cgctx);
CGDataProviderRef provider = CGDataProviderCreateWithData(NULL, bufferdata, mWidth*mHeight*4, NULL);
CGImageRef savedimageref = CGImageCreate(mWidth,mHeight, 8, 32, mWidth*4, colorSpaceRef, bitmapInfo,provider , NULL, NO, renderingIntent);
CFRelease(colorSpaceRef);
return savedimageref;
}
then call changecolor method on a buttons click event and see the result
I found the answer for my above problem, Follow the above code of Rahul with some changes to set your own color,
-(void)changeColor
{
UIImage *temp23=Image;//Pass your image here
CGImageRef ref1=[self createMask:temp23];
const float colorMasking[6] = {1.0, 3.0, 1.0, 2.0, 2.0, 3.0};
CGImageRef New=CGImageCreateWithMaskingColors(ref1, colorMasking);
UIImage *resultedimage=[UIImage imageWithCGImage:New];
EditImageView.image = resultedimage;
[EditImageView setNeedsDisplay];
}
-(CGImageRef)createMask:(UIImage*)temp
{
CGImageRef ref=temp.CGImage;
int mWidth=CGImageGetWidth(ref);
int mHeight=CGImageGetHeight(ref);
int count=mWidth*mHeight*4;
void *bufferdata=malloc(count);
CGColorSpaceRef colorSpaceRef = CGColorSpaceCreateDeviceRGB();
CGBitmapInfo bitmapInfo = kCGBitmapByteOrderDefault;
CGColorRenderingIntent renderingIntent = kCGRenderingIntentDefault;
CGContextRef cgctx = CGBitmapContextCreate (bufferdata,mWidth,mHeight, 8,mWidth*4,colorSpaceRef, kCGImageAlphaPremultipliedLast);
CGRect rect = {0,0,mWidth,mHeight};
CGContextDrawImage(cgctx, rect, ref);
CGContextSaveGState(cgctx);
CGContextSetBlendMode(cgctx, kCGBlendModeColor);
CGContextSetRGBFillColor (cgctx, 1.0, 0.0, 0.0, 1.0);
CGContextFillRect(cgctx, rect);
bufferdata = CGBitmapContextGetData (cgctx);
const float colorMasking[6] = {1.0, 3.0, 1.0, 2.0, 2.0, 3.0};
CGDataProviderRef provider = CGDataProviderCreateWithData(NULL, bufferdata, mWidth*mHeight*4, NULL);
CGImageRef savedimageref = CGImageCreate(mWidth,mHeight, 8, 32, mWidth*4, colorSpaceRef, bitmapInfo,provider , NULL, NO, renderingIntent);
CFRelease(colorSpaceRef);
return savedimageref;
}
Hmmm...
I may be missing something, but I don't believe that the provided answers apply to the question. The second response gets closer to the mark, but it contains spurious code that has little to do with the solution.
The createMask method makes a copy of the original image assuming alpha in LSB position. The changeColor performs a masking call that isn't going to do much to an RGB image -- basically only black is going to be masked (i.e, RGB triplets in the range/combinations of (1,1,2) to (3,2,1)).
I am guessing that the red shift that is being observed is due to the parameter
kCGImageAlphaPremultipliedFirst
in the line
CGContextRef cgctx = CGBitmapContextCreate (bufferdata,mWidth,mHeight, 8,mWidth*4, colorSpaceRef, kCGImageAlphaPremultipliedFirst);
due to improper treatment of the alpha channel. If in the changeColor method you modify the block
CGImageRef ref1=[self createMask:temp23];
const float colorMasking[6] = {1.0, 3.0, 1.0, 2.0, 2.0, 3.0};
CGImageRef New=CGImageCreateWithMaskingColors(ref1, colorMasking);
UIImage *resultedimage=[UIImage imageWithCGImage:New];
EditImageView.image = resultedimage;
to be
CGImageRef ref1=[self createMask:temp23];
UIImage *resultedimage=[UIImage imageWithCGImage:ref];
EditImageView.image = resultedimage;
you'll see no difference in the display. Changing the CGBitmapInfo constant to kCGImageAlphaPremultipliedLast should display the image correctly using either of the above code blocks.
The next response gets a bit closer to what the OP asks for, but its in terms of visual effect, not actual data. Here the pertinent code in createMask is
CGContextRef cgctx = CGBitmapContextCreate (bufferdata,mWidth,mHeight, 8,mWidth*4,colorSpaceRef, kCGImageAlphaPremultipliedLast);
which displays the image correctly, followed by
CGContextSetBlendMode(cgctx, kCGBlendModeColor);
CGContextSetRGBFillColor (cgctx, 1.0, 0.0, 0.0, 1.0);
CGContextFillRect(cgctx, rect);
and then the logic for constructing the image. The blend logic overlays a red tint on the original image, achieving a similar effect as the misplaced alpha channel in the original response. This still really is not what the OP asks for, which is to mask one or more channels, not blend colors.
This really amounts to setting the channel values for the colors that are not desired to zero. Here's an example for returning just the red channel as the OP requests; the assumption is that the format of the pixel is ABGR:
- (CGImageRef) redChannel:(CGImageRef)image
{
CGDataProviderRef provider = CGImageGetDataProvider(image);
NSMutableData* data = (id)CGDataProviderCopyData(provider);
int width = CGImageGetWidth(image);
int height = CGImageGetHeight(image);
[data autorelease];
// get a mutable reference to the image data
uint32_t* dwords = [data mutableBytes];
for (size_t idx = 0; idx < width*height; idx++) {
uint32_t* pixel = &dwords[idx];
// perform a logical AND of the pixel with a mask that zeroes out green and blue pixels
*pixel &= 0x000000ff;
}
CGColorSpaceRef colorSpaceRef = CGColorSpaceCreateDeviceRGB();
CGBitmapInfo bitmapInfo = kCGBitmapByteOrderDefault;
CGColorRenderingIntent renderingIntent = kCGRenderingIntentDefault;
// now create a new image using the masked original data
CGDataProviderRef iprovider = CGDataProviderCreateWithData(NULL, dwords, width*height*4, NULL);
CGImageRef savedimageref = CGImageCreate(width, height, 8, 32, width*4, colorSpaceRef, bitmapInfo, iprovider, NULL, NO, renderingIntent);
CGColorSpaceRelease(colorSpaceRef);
CGDataProviderRelease(iprovider);
return savedimageref;
}
A good summary of bitwise operations can be found here.
As is pointed out here, you may need to change the structure of the mask depending on the LSB/MSB ordering of the bits in the pixel. This example assumes 32 bit pixels from a standard iPhone PNG.