I've got a UIImageView (full frame and rectangular) that i'm rotating with a CGAffineTransform. The UIImage of the UIImageView fills the entire frame. When the image is rotated and drawn the edges appear noticeably jagged. Is there anything I can do to make it look better? It's clearly not being anti-aliased with the background.
The edges of CoreAnimation layers aren't antialiased by default on iOS. However, there is a key that you can set in Info.plist that enables antialiasing of the edges: UIViewEdgeAntialiasing.
https://developer.apple.com/library/content/documentation/General/Reference/InfoPlistKeyReference/Articles/iPhoneOSKeys.html
If you don't want the performance overhead of enabling this option, a work-around is to add a 1px transparent border around the edge of the image. This means that the 'edges' of the image are no longer on the edge, so don't need special treatment!
New API – iOS 6/7
Also works for iOS 6, as noted by #Chris, but wasn't made public until iOS 7.
Since iOS 7, CALayer has a new property allowsEdgeAntialiasing which does exactly what you want in this case, without incurring the overhead of enabling it for all views in your application! This is a property of CALayer, so to enable this for a UIView you use myView.layer.allowsEdgeAntialiasing = YES.
just add 1px transparent border to your image
CGRect imageRect = CGRectMake(0, 0, image.size.width, image.size.height);
UIGraphicsBeginImageContextWithOptions(imageRect.size, NO, 0.0);
[image drawInRect:CGRectMake(1,1,image.size.width-2,image.size.height-2)];
image = UIGraphicsGetImageFromCurrentImageContext();
UIGraphicsEndImageContext();
Remember to set the appropriate anti-alias options:
CGContextSetAllowsAntialiasing(theContext, true);
CGContextSetShouldAntialias(theContext, true);
just add "Renders with edge antialiasing" with YES in plist and it will work.
I would totally recommend the following library.
http://vocaro.com/trevor/blog/2009/10/12/resize-a-uiimage-the-right-way/
It contains lots of useful extensions to UIImage that solve this problem and also include code for generating thumbnails etc.
Enjoy!
The best way I've found to have smooth edges and a sharp image is to do this:
CGRect imageRect = CGRectMake(0, 0, self.photo.image.size.width, self.photo.image.size.height);
UIGraphicsBeginImageContextWithOptions(imageRect.size, NO, 0.0);
[self.photo.image drawInRect:CGRectMake(1, 1, self.photo.image.size.width - 2, self.photo.image.size.height - 2)];
self.photo.image = UIGraphicsGetImageFromCurrentImageContext();
UIGraphicsEndImageContext();
Adding the Info.plist key like some people describe has a big hit on performance and if you use that then you're basically applying it to everything instead of just the one place you need it.
Also, don't just use UIGraphicsBeginImageContext(imageRect.size); otherwise the layer will be blurry. You have to use UIGraphicsBeginImageContextWithOptions like I've shown.
I found this solution from here, and it's perfect:
+ (UIImage *)renderImageFromView:(UIView *)view withRect:(CGRect)frame transparentInsets:(UIEdgeInsets)insets {
CGSize imageSizeWithBorder = CGSizeMake(frame.size.width + insets.left + insets.right, frame.size.height + insets.top + insets.bottom);
// Create a new context of the desired size to render the image
UIGraphicsBeginImageContextWithOptions(imageSizeWithBorder, NO, 0);
CGContextRef context = UIGraphicsGetCurrentContext();
// Clip the context to the portion of the view we will draw
CGContextClipToRect(context, (CGRect){{insets.left, insets.top}, frame.size});
// Translate it, to the desired position
CGContextTranslateCTM(context, -frame.origin.x + insets.left, -frame.origin.y + insets.top);
// Render the view as image
[view.layer renderInContext:UIGraphicsGetCurrentContext()];
// Fetch the image
UIImage *renderedImage = UIGraphicsGetImageFromCurrentImageContext();
// Cleanup
UIGraphicsEndImageContext();
return renderedImage;
}
usage:
UIImage *image = [UIImage renderImageFromView:view withRect:view.bounds transparentInsets:UIEdgeInsetsZero];
Related
I'm rendering PDF content into a UIView and I'm seeing that the text provided by the PDF is blurry when zoomed.
The way I'm rendering the text is as follows
CGSize size = CGSizeMake(96, 9); // These numbers come from the PDF
NSString* text = #"Text to render";
UIGraphicsBeginImageContextWithOptions(size, NO, 0.0);
[text drawInRect:CGRectMake(0, 0, size.width, size.height)];
UIImage *image = UIGraphicsGetImageFromCurrentImageContext();
UIGraphicsEndImageContext();
This UIImage has the correct size and when I inspect it via XCode it is crisp.
When I call CGContextDrawImage below
CGRect widgetRect = CGRectMake(0,0,180,90);
CGContextDrawImage(mainContext, size, image.CGImage);
The result is blurry.
Notes:
The mainContext above has it's origin lower left so that's why I render text into a separate context and draw an image.
The UIView has a contentScaleFactor of 3 and the mainContext has a size to match that scale.
I looked at CGContextDrawImage draws large images very blurry and it doesn't address my problem.
I can't reproduce this problem outside my app in it's own.
This last part shows me that the problem is somewhere in the app code so I' hoping for hints and ideas about where to look in the rendering pipeline and how to debug.
EDIT:
Updated the calll to CGContextDrawImage to use the correct size.
I've been having issues rendering images with the UIImageView class. The pixelation seems to occur mostly on the edges of the image I am trying to show.
I have tried changing the property 'Render with edge antialiasing' to no avail.
The image files contain images that are larger than what will appear on the screen.
It seems to be royally messing with the quality of the image and then displaying it. I tried to post images here, but StackOverflow is denying me that privilege. So here's a link to what's going on.
http://i.imgur.com/QpUOTOF.png
The sun in this image is the problem I'm speaking of. Any ideas?
On-the-fly image resizing is quick and of low quality. For bundled images, it is worth the extra bundle space to include downsized versions. For downloaded images, you can achieve better results by resizing with Core Graphics into a new UIImage before you set the image property.
CGSize newSize = CGSizeMake(newWidth, newHeight);
UIGraphicsBeginImageContextWithOptions(newSize, // context size
NO, // opaque?
0); // image scale. 0 means "device screen scale"
CGContextRef context = UIGraphicsGetCurrentContext();
CGContextSetInterpolationQuality(context, kCGInterpolationHigh);
[bigImage drawInRect:CGRectMake(0, 0, newSize.width, newSize.height)];
UIImage *newImage = UIGraphicsGetImageFromCurrentImageContext();
UIGraphicsEndImageContext();
Use following method use for get specific hight and width with image
+ (UIImage*)resizeImage:(UIImage*)image withWidth:(int)width withHeight:(int)height
{
CGSize newSize = CGSizeMake(width, height);
float widthRatio = newSize.width/image.size.width;
float heightRatio = newSize.height/image.size.height;
if(widthRatio > heightRatio)
{
newSize=CGSizeMake(image.size.width*heightRatio,image.size.height*heightRatio);
}
else
{
newSize=CGSizeMake(image.size.width*widthRatio,image.size.height*widthRatio);
}
UIGraphicsBeginImageContextWithOptions(newSize, NO, 0.0);
[image drawInRect:CGRectMake(0,0,newSize.width,newSize.height)];
UIImage* newImage = UIGraphicsGetImageFromCurrentImageContext();
UIGraphicsEndImageContext();
return newImage;
}
This method return NewImage, with specific size that you specified.
How big is your image and what is the size of the imageView? Don't rely on UIImageView to scale it down for you. You probably need to resize it manually. This would also be a bit more memory efficient.
I use categories like these:
>>>github link <<<
to do image resizing.
This also gives you some other nice function for rounded corners etc.
Also keep in mind, that you need a transparent border at the edge of an image if you want to rotate it to avoid aliasing.
I have a view (called outPutView) that contains graphics, like uIImageViews and labels. I need to render an image of the outPutView and it's sub-views. I am using renderInContext:UIGraphicsGetCurrentContext() to do so. Works fine, except that I need to scale the views. I am using a transform on the outPutView. This successfully scales the view and it's sub-views, but the transform does not render. While the views are scaled onscreen. the final render displays the vies at their original size, while the render context is at the target size (here #2x iPhone view size).
Thanks for reading!!
[outPutView setTransform:CGAffineTransformMake(2, 0, 0, 2, 0, 0)];
CGSize renderSize = CGSizeMake(self.view.bounds.size.width*2, self.view.bounds.size.height*2);
UIGraphicsBeginImageContext(renderSize);
[[outPutView layer] renderInContext:UIGraphicsGetCurrentContext()];
renderedImage = UIGraphicsGetImageFromCurrentImageContext();
UIGraphicsEndImageContext();
I've just made this work, although in the opposite direction (scaling down). Here's a summary of the relevant code:
// destination size is half of self (self is a UIView)
float rescale = 0.5;
CGSize resize = CGSizeMake(self.width * rescale, self.height * rescale);
// make the destination context
UIGraphicsBeginImageContextWithOptions(resize, YES, 0);
// apply the scale to dest context
CGContextScaleCTM(UIGraphicsGetCurrentContext(), rescale, rescale);
// render self into dest context
[self.layer renderInContext:UIGraphicsGetCurrentContext()];
// grab the resulting UIImage
UIImage *image = UIGraphicsGetImageFromCurrentImageContext();
UIGraphicsEndImageContext();
To scale up instead of down, it should be fine to have rescale = 2.
I solved this by re-ordering my views. Actually adding another view between the output view: the view that the rendered context is taken from, and the view that is scaled via transform. It worked, but I have no idea why at this point. Any thoughts on this would be appreciated. Thanks for reading.
I have an image inside an UIImageView which is within a UIScrollView. What I want to do is rotate this image 90 degrees so that it is in landscape by default, and set the initial zoom of the image so that the entire image fits into the scrollview and then allow it to be zoomed up to 100% and back down to minimum zoom again.
This is what I have so far:
self.imageView.transform = CGAffineTransformMakeRotation(-M_PI/2);
float minimumScale = scrollView.frame.size.width / self.imageView.frame.size.width;
scrollView.minimumZoomScale = minimumScale;
scrollView.zoomScale = minimumScale;
scrollView.contentSize = CGSizeMake(self.imageView.frame.size.height,self.imageView.frame.size.width);
The problem is that if I set the transform, nothing shows up in the scrollview. However if I commented out the transform, everything works except the image is not in the landscape orientation that I want it to be!
If I apply the transform and remove the code that sets the minimumZoomScale and zoomScale properties, then the image shows up in the correct orientation, however with the incorrect zoomScale and seems like the contentSize property isn't set correctly either - since the doesn't scroll to the edge of the image in the left/right direction, however does top and bottom but much over the edge.
NB: image is being loaded from a URL
Maybe rotating the image itself fits your needs:
UIImage* rotateUIImage(const UIImage* src, float angleDegrees) {
UIView* rotatedViewBox = [[UIView alloc] initWithFrame: CGRectMake(0, 0, src.size.width, src.size.height)];
float angleRadians = angleDegrees * ((float)M_PI / 180.0f);
CGAffineTransform t = CGAffineTransformMakeRotation(angleRadians);
rotatedViewBox.transform = t;
CGSize rotatedSize = rotatedViewBox.frame.size;
[rotatedViewBox release];
UIGraphicsBeginImageContext(rotatedSize);
CGContextRef bitmap = UIGraphicsGetCurrentContext();
CGContextTranslateCTM(bitmap, rotatedSize.width/2, rotatedSize.height/2);
CGContextRotateCTM(bitmap, angleRadians);
CGContextScaleCTM(bitmap, 1.0, -1.0);
CGContextDrawImage(bitmap, CGRectMake(-src.size.width / 2, -src.size.height / 2, src.size.width, src.size.height), [src CGImage]);
UIImage *newImage = UIGraphicsGetImageFromCurrentImageContext();
UIGraphicsEndImageContext();
return newImage;
}
I believe the easiest way (and thread safe too) is to do:
//assume that the image is loaded in landscape mode from disk
UIImage * LandscapeImage = [UIImage imageNamed: imgname];
UIImage * PortraitImage = [[UIImage alloc] initWithCGImage: LandscapeImage.CGImage
scale: 1.0
orientation: UIImageOrientationLeft];
Any calculations that you do based on the imageView's frame should probably be done before you apply any transformations to it. But I would actually suggest doing those calculations based on the size of the UIImage, not the UIImageView. Then set both the UIImageView's frame and the UIScrollView's contentSize based on that.
Max's suggestion is a good one, although with a larger image it could be a performance killer. Are you displaying this image from your app's resources? If so, why not just rotate the images before you even build the app?
There's a much easier solution that is also faster, just do this:
- (void) imageRotateTapped:(id)sender
{
[UIView animateWithDuration:0.33f animations:^()
{
self.imageView.transform = CGAffineTransformMakeRotation(RADIANS(self.rotateDegrees += 90.0f));
self.imageView.frame = self.imageView.superview.bounds; // change this to whatever rect you want
}];
}
When the user is done, you will need to actually create a new rotated image, but that is very easy to do.
I was using the accepted answer for a while until we noticed that non-square rotations based on images taken directly from the camera seemed stretched (they were rotated as desired, just the frame width/height wasn't adjusted).
Great explanation/post here from Trevor: http://vocaro.com/trevor/blog/2009/10/12/resize-a-uiimage-the-right-way/
In the end, it was a very simple import of Trevor's code which uses categories to add a resizedImage:interpoationQuality method to UIImage. So yeah, user beware, if it still works for you, great. But if it doesn't, I'd take a look at the library instead.
I like the way this (http://shakeitphoto.com/) application puts a border around the image.. i would like to do something similar in my application but not sure how should I go about doing it.
Any ideas on how given a UIImage can I wrap a frame around it?
From that website, it appears you want a border with a shadow. There's 2 reasonable options, 3 if you don't care about the shadow.
If you don't care about the shadow, you can just do something like
#import <QuartzCore/QuartzCore.h> // this should be at the top
// inside your view layout code
myImageView.layer.borderColor = [UIColor whiteColor].CGColor
myImageView.layer.borderWidth = 5;
This will give you a 5-pixel white border inset into the view, layered on top of the view's contents (e.g. the image). What it won't give you is a shadow. If you want the shadow, there's 2 other options.
You could just create an image that includes the border and the shadow, and nothing else. Just make everything else alpha-transparent. Then you can simply layer this image on top of the one you want to display (either with 2 imageviews, or by creating a third image out of the 2). This should work fine, but it won't scale to different image sizes. In the case of the linked app, the image size is always the same so they could be using this.
The other option is to simply draw the border and shadow on top of your image in a new image. Here's a bit of sample code that will do this - it creates a new image the same size as your original, but with a white, shadowed border:
- (UIImage *)borderedImage:(UIImage *)image {
// the following NO means the new image has an alpha channel
// If you know the source image is fully-opaque, you may want to set that to YES
UIGraphicsBeginImageContextWithOptions(image.size, NO, image.scale);
[image drawAtPoint:CGPointZero];
CGContextRef ctx = UIGraphicsGetCurrentContext();
const CGFloat shadowRadius = 5;
CGContextSetShadowWithColor(ctx, 0, shadowRadius, [UIColor blackColor].CGColor);
[[UIColor whiteColor] set];
CGRect rect = (CGRect){CGPointZero, image.size};
const CGFloat frameWidth = 5;
rect = CGRectInset(rect, frameWidth / 2.0f, frameWidth / 2.0f);
UIBezierPath *path = [UIBezierPath bezierPathWithRect:rect];
path.lineWidth = frameWidth;
[path stroke];
UIImage *newImage = UIGraphicsGetImageFromCurrentImageContext();
// note: getting the new image this way throws away the orientation data from the original
// You could create a third image by doing something like
// newImage = [UIImage imageWithCGImage:newImage.CGImage scale:newImage.scale orientation:image.orientation]
// but I am unsure as to how orientation actually affects rendering (if at all)
UIGraphicsEndImageContext();
return newImage;
}
(note: this code has not been compiled and could contain bugs)