Translate detected rectangle from portrait CIImage to landscape CIImage - objective-c

I am modifying code found here. In the code we are capturing video from the phone camera using AVCaptureSession and using CIDetector to detect a rectangle in the image feed. The feed has an image which is 640x842 (iphone5 in portrait). We then do an overlay on the image so the user can see the detected rectangle (actually it's a trapezoid most of the time).
When the user presses a button on the UI, we capture an image from the video and re-run the rectangle detection on this larger image (3264x2448) which as you can see is landscape. We then do a perspective transform on the detected rectangle and crop on the image.
This is working pretty well but the issue I have is say 1 out of 5 captures the detected rectangle on the larger image is different to the one detected (and presented to the user) from the smaller image. Even though I only capture when I detect the phone is (relatively) still, the final image then does not represent the rectangle the user expected.
To resolve this my idea is to use the coordinates of the originally captured rectangle and translate them to a rectangle on the captured still image. This is where I'm struggling.
I tried this with the detected rectangle:
CGFloat radians = -90 * (M_PI/180);
CGAffineTransform rotation = CGAffineTransformMakeRotation(radians);
CGRect rect = CGRectMake(detectedRect.bounds.origin.x, detectedRect.bounds.origin.y, detectedRect.bounds.size.width, detectedRect.bounds.size.height);
CGRect rotatedRect = CGRectApplyAffineTransform(rect, rotation);
So given a detected rect:
TopLeft: 88.213425, 632.31329
TopRight: 545.59302, 632.15546
BottomRight: 575.57819, 369.22321
BottomLeft: 49.973862, 369.40466
I now have this rotated rect:
origin = (x = 369.223206, y = -575.578186)
size = (width = 263.090088, height = 525.604309)
How do I translate the rotated rectangle coordinates in the smaller portrait image to coordinates to the 3264x2448 image?
Edit
Duh.. reading my own approach realised that creating a rectangle out of a trapezoid will not solve my problem!
Supporting code to detect the rectangle etc...
// In this method we detect a rect from the video feed and overlay
-(void)captureOutput:(AVCaptureOutput *)captureOutput didOutputSampleBuffer:(CMSampleBufferRef)sampleBuffer fromConnection:(AVCaptureConnection *)connection
{
CVPixelBufferRef pixelBuffer = (CVPixelBufferRef)CMSampleBufferGetImageBuffer(sampleBuffer);
CIImage *image = [CIImage imageWithCVPixelBuffer:pixelBuffer];
// image is 640x852 on iphone5
NSArray *rects = [[CIDetector detectorOfType:CIDetectorTypeRectangle context:nil options:#{CIDetectorAccuracy : CIDetectorAccuracyHigh}] featuresInImage:image];
CIRectangleFeature *detectedRect = rects[0];
// draw overlay on image code....
}
This is a summarized version of how the still image is obtained:
// code block to handle output from AVCaptureStillImageOutput
[self.stillImageOutput captureStillImageAsynchronouslyFromConnection:videoConnection completionHandler: ^(CMSampleBufferRef imageSampleBuffer, NSError *error)
{
NSData *imageData = [AVCaptureStillImageOutput jpegStillImageNSDataRepresentation:imageSampleBuffer];
CIImage *enhancedImage = [[CIImage alloc] initWithData:imageData options:#{kCIImageColorSpace:[NSNull null]}];
imageData = nil;
CIRectangleFeature *rectangleFeature = [self getDetectedRect:[[self highAccuracyRectangleDetector] featuresInImage:enhancedImage]];
if (rectangleFeature) {
enhancedImage = [self correctPerspectiveForImage:enhancedImage withTopLeft:rectangleFeature.topLeft andTopRight:rectangleFeature.topRight andBottomRight:rectangleFeature.bottomRight andBottomLeft:rectangleFeature.bottomLeft];
}
}
Thank you.

I had the same issue while doing this kind of stuff.I have resolved it by doing below code in swift. Take a look if it can help you.
if let videoConnection = stillImageOutput.connection(withMediaType: AVMediaTypeVideo) {
stillImageOutput.captureStillImageAsynchronously(from: videoConnection) {
(imageDataSampleBuffer, error) -> Void in
let imageData = AVCaptureStillImageOutput.jpegStillImageNSDataRepresentation(imageDataSampleBuffer)
var img = UIImage(data: imageData!)!
let outputRect = self.previewLayer?.metadataOutputRectOfInterest(for: (self.previewLayer?.bounds)!)
let takenCGImage = img.cgImage
let width = (takenCGImage?.width)!
let height = (takenCGImage?.height)!
let cropRect = CGRect(x: (outputRect?.origin.x)! * CGFloat(width), y: (outputRect?.origin.y)! * CGFloat(height), width: (outputRect?.size.width)! * CGFloat(width), height: (outputRect?.size.height)! * CGFloat(height))
let cropCGImage = takenCGImage!.cropping(to: cropRect)
img = UIImage(cgImage: cropCGImage!, scale: 1, orientation: img.imageOrientation)
let cropViewController = TOCropViewController(image: self.cropToBounds(image: img))
cropViewController.delegate = self
self.navigationController?.pushViewController(cropViewController, animated: true)
}
}

Related

How Do I Blur a Scene in SpriteKit?

How would I add a gaussian blur to all nodes (there's no fixed number of nodes) in an SKScene in SpriteKit? A label will be added on top of the scene later, this will be my pause menu.
Almost anything would help!
Something like this is what I'm going for:
What you're looking for is an SKEffectNode. It applies a CoreImage filter to itself (and thus all subnodes). Just make it the root view of your scene, give it one of CoreImage's blur filters, and you're set.
For example, I set up an SKScene with an SKEffectNode as it's first child node and a property, root that holds a weak reference to it:
-(void)createLayers{
SKEffectNode *node = [SKEffectNode node];
[node setShouldEnableEffects:NO];
CIFilter *blur = [CIFilter filterWithName:#"CIGaussianBlur" keysAndValues:#"inputRadius", #1.0f, nil];
[node setFilter:blur];
[self setRoot:node];
}
And here's the method I use to (animate!) the blur of my scene:
-(void)blurWithCompletion:(void (^)())handler{
CGFloat duration = 0.5f;
[[self root] setShouldRasterize:YES];
[[self root] setShouldEnableEffects:YES];
[[self root] runAction:[SKAction customActionWithDuration:duration actionBlock:^(SKNode *node, CGFloat elapsedTime){
NSNumber *radius = [NSNumber numberWithFloat:(elapsedTime/duration) * 10.0];
[[(SKEffectNode *)node filter] setValue:radius forKey:#"inputRadius"];
}] completion:handler];
}
Note that, like you, I'm using this as a pause screen, so I rasterize the scene. If you want your scene to animate while blurred, you should probably setShouldResterize: to NO.
And if you're not interested in animating the transition to the blur, you could always just set the filter to an initial radius of 10.0f or so and do a simple setShouldEnableEffects:YES when you want to switch it on.
See also: SKEffectNode class reference
UPDATE:
See Markus's comment below. He points out that SKScene is, in fact, a subclass of SKEffectNode, so you really ought to be able to call all of this on the scene itself rather than arbitrarily inserting an effect node in your node tree.
To add to this by using #Bendegúz's answer and code from http://www.bytearray.org/?p=5360
I was able to get this to work in my current game project that's being done in IOS 8 Swift. Done a bit differently by returning an SKSpriteNode instead of a UIImage. Also note that my unwrapped currentScene.view! call is to a weak GameScene reference but should work with self.view.frame based on where you are calling these methods. My pause screen is called in a separate HUD class hence why this is the case.
I would imagine this could be done more elegantly, maybe more like #jemmons's answer. Just wanted to possibly help out anyone else trying to do this in SpriteKit projects written in all or some Swift code.
func getBluredScreenshot() -> SKSpriteNode{
create the graphics context
UIGraphicsBeginImageContextWithOptions(CGSize(width: currentScene.view!.frame.size.width, height: currentScene.view!.frame.size.height), true, 1)
currentScene.view!.drawViewHierarchyInRect(currentScene.view!.frame, afterScreenUpdates: true)
// retrieve graphics context
let context = UIGraphicsGetCurrentContext()
// query image from it
let image = UIGraphicsGetImageFromCurrentImageContext()
// create Core Image context
let ciContext = CIContext(options: nil)
// create a CIImage, think of a CIImage as image data for processing, nothing is displayed or can be displayed at this point
let coreImage = CIImage(image: image)
// pick the filter we want
let filter = CIFilter(name: "CIGaussianBlur")
// pass our image as input
filter.setValue(coreImage, forKey: kCIInputImageKey)
//edit the amount of blur
filter.setValue(3, forKey: kCIInputRadiusKey)
//retrieve the processed image
let filteredImageData = filter.valueForKey(kCIOutputImageKey) as CIImage
// return a Quartz image from the Core Image context
let filteredImageRef = ciContext.createCGImage(filteredImageData, fromRect: filteredImageData.extent())
// final UIImage
let filteredImage = UIImage(CGImage: filteredImageRef)
// create a texture, pass the UIImage
let texture = SKTexture(image: filteredImage!)
// wrap it inside a sprite node
let sprite = SKSpriteNode(texture:texture)
// make image the position in the center
sprite.position = CGPointMake(CGRectGetMidX(currentScene.frame), CGRectGetMidY(currentScene.frame))
var scale:CGFloat = UIScreen.mainScreen().scale
sprite.size.width *= scale
sprite.size.height *= scale
return sprite
}
func loadPauseBGScreen(){
let duration = 1.0
let pauseBG:SKSpriteNode = self.getBluredScreenshot()
//pauseBG.position = CGPointMake(CGRectGetMidX(self.frame), CGRectGetMidY(self.frame))
pauseBG.alpha = 0
pauseBG.zPosition = self.zPosition + 1
pauseBG.runAction(SKAction.fadeAlphaTo(1, duration: duration))
self.addChild(pauseBG)
}
This is my solution for the pause screen.
It will take a screenshot, blur it and after that show it with animation.
I think you should do it if you don't wanna waste to much fps.
-(void)pause {
SKSpriteNode *pauseBG = [SKSpriteNode spriteNodeWithTexture:[SKTexture textureWithImage:[self getBluredScreenshot]]];
pauseBG.position = CGPointMake(CGRectGetMidX(self.frame), CGRectGetMidY(self.frame));
pauseBG.alpha = 0;
pauseBG.zPosition = 2;
[pauseBG runAction:[SKAction fadeAlphaTo:1 duration:duration / 2]];
[self addChild:pauseBG];
}
And this is the helper method:
- (UIImage *)getBluredScreenshot {
UIGraphicsBeginImageContextWithOptions(self.view.bounds.size, NO, 1);
[self.view drawViewHierarchyInRect:self.view.frame afterScreenUpdates:YES];
UIImage *ss = UIGraphicsGetImageFromCurrentImageContext();
UIGraphicsEndImageContext();
CIFilter *gaussianBlurFilter = [CIFilter filterWithName:#"CIGaussianBlur"];
[gaussianBlurFilter setDefaults];
[gaussianBlurFilter setValue:[CIImage imageWithCGImage:[ss CGImage]] forKey:kCIInputImageKey];
[gaussianBlurFilter setValue:#10 forKey:kCIInputRadiusKey];
CIImage *outputImage = [gaussianBlurFilter outputImage];
CIContext *context = [CIContext contextWithOptions:nil];
CGRect rect = [outputImage extent];
rect.origin.x += (rect.size.width - ss.size.width ) / 2;
rect.origin.y += (rect.size.height - ss.size.height) / 2;
rect.size = ss.size;
CGImageRef cgimg = [context createCGImage:outputImage fromRect:rect];
UIImage *image = [UIImage imageWithCGImage:cgimg];
CGImageRelease(cgimg);
return image;
}
Swift 4:
add this to your gameScene if you want to blur everything in the scene:
let blur = CIFilter(name:"CIGaussianBlur",withInputParameters: ["inputRadius": 10.0])
self.filter = blur
self.shouldRasterize = true
self.shouldEnableEffects = false
change self.shouldEnableEffects = true when you want to use it.
This is another example of getting this done in swift 2 without the layers:
func blurWithCompletion() {
let duration: CGFloat = 0.5
let filter: CIFilter = CIFilter(name: "CIGaussianBlur", withInputParameters: ["inputRadius" : NSNumber(double:1.0)])!
scene!.filter = filter
scene!.shouldRasterize = true
scene!.shouldEnableEffects = true
scene!.runAction(SKAction.customActionWithDuration(0.5, actionBlock: { (node: SKNode, elapsedTime: CGFloat) in
let radius = (elapsedTime/duration)*10.0
(node as? SKEffectNode)!.filter!.setValue(radius, forKey: "inputRadius")
}))
}
Swift 3 Update: This is #Chuck Gaffney's answer updated for Swift 3. I know this question is tagged objective-c, but this page ranked 2nd in Google for "swift spritekit blur". I changed currentScene to self.
func getBluredScreenshot() -> SKSpriteNode{
//create the graphics context
UIGraphicsBeginImageContextWithOptions(CGSize(width: self.view!.frame.size.width, height: self.view!.frame.size.height), true, 1)
self.view!.drawHierarchy(in: self.view!.frame, afterScreenUpdates: true)
// retrieve graphics context
_ = UIGraphicsGetCurrentContext()
// query image from it
let image = UIGraphicsGetImageFromCurrentImageContext()
// create Core Image context
let ciContext = CIContext(options: nil)
// create a CIImage, think of a CIImage as image data for processing, nothing is displayed or can be displayed at this point
let coreImage = CIImage(image: image!)
// pick the filter we want
let filter = CIFilter(name: "CIGaussianBlur")
// pass our image as input
filter?.setValue(coreImage, forKey: kCIInputImageKey)
//edit the amount of blur
filter?.setValue(3, forKey: kCIInputRadiusKey)
//retrieve the processed image
let filteredImageData = filter?.value(forKey: kCIOutputImageKey) as! CIImage
// return a Quartz image from the Core Image context
let filteredImageRef = ciContext.createCGImage(filteredImageData, from: filteredImageData.extent)
// final UIImage
let filteredImage = UIImage(cgImage: filteredImageRef!)
// create a texture, pass the UIImage
let texture = SKTexture(image: filteredImage)
// wrap it inside a sprite node
let sprite = SKSpriteNode(texture:texture)
// make image the position in the center
sprite.position = CGPoint(x: self.frame.midX, y: self.frame.midY)
let scale:CGFloat = UIScreen.main.scale
sprite.size.width *= scale
sprite.size.height *= scale
return sprite
}
func loadPauseBGScreen(){
let duration = 1.0
let pauseBG:SKSpriteNode = self.getBluredScreenshot()
pauseBG.alpha = 0
pauseBG.zPosition = self.zPosition + 1
pauseBG.run(SKAction.fadeAlpha(to: 1, duration: duration))
self.addChild(pauseBG)
}
I was trying to do this same thing now, in May 2020 (Xcode 11 and iOS 13.x), but wasn't unable to 'animate' the blur radius. In my case, I start with the scene fully blurred, and then 'unblur' it gradually (set inputRadius to 0).
Somehow, the new input radius value set in the custom action block wasn't reflected in the rendered scene. My code was as follows:
private func unblur() {
run(SKAction.customAction(withDuration: unblurDuration, actionBlock: { [weak self] (_, elapsed) in
guard let this = self else { return }
let ratio = (TimeInterval(elapsed) / this.unblurDuration)
let radius = this.maxBlurRadius * (1 - ratio) // goes to 0 as ratio goes to 1
this.filter?.setValue(radius, forKey: kCIInputRadiusKey)
}))
}
I even tried updating the value manually using SKScene.update(_:) and some variables for time book-keeping, but the same result.
It occurred to me that perhaps I could force the refresh if I "re-assingned" the blur filter to the .filter property of my SKScene (see comments in ALL CAPS near the end of the code), to the same effect, and it worked.
The full code:
class MyScene: SKScene {
private let maxBlurRadius: Double = 50
private let unblurDuration: TimeInterval = 5
init(size: CGSize) {
super.init(size: size)
let filter = CIFilter(name: "CIGaussianBlur")
filter?.setValue(maxBlurRadius, forKey: kCIInputRadiusKey)
self.filter = filter
self.shouldEnableEffects = true
self.shouldRasterize = false
// (...rest of the child nodes, etc...)
}
override func didMove(to view: SKView) {
super.didMove(to: view)
self.unblur()
}
private func unblur() {
run(SKAction.customAction(withDuration: unblurDuration, actionBlock: { [weak self] (_, elapsed) in
guard let this = self else { return }
let ratio = (TimeInterval(elapsed) / this.unblurDuration)
let radius = this.maxBlurRadius * (1 - ratio) // goes to 0 as ratio goes to 1
// OBTAIN THE FILTER
let filter = this.filter
// MODIFY ATTRIBUTE
filter?.setValue(radius, forKey: kCIInputRadiusKey)
// RE=ASSIGN TO SCENE
this.filter = filter
}))
}
}
I hope this helps someone!

Get the correct image width and height of an NSImage

I use the code below to get the width and height of a NSImage:
NSImage *image = [[[NSImage alloc] initWithContentsOfFile:[NSString stringWithFormat:s]] autorelease];
imageWidth=[image size].width;
imageHeight=[image size].height;
NSLog(#"%f:%f",imageWidth,imageHeight);
But sometime imageWidth, imageHeight does not return the correct value. For example when I read an image, the EXIF info displays:
PixelXDimension = 2272;
PixelYDimension = 1704;
But imageWidth, imageHeight outputs
521:390
Dimensions of your image in pixels is stored in NSImageRep of your image. If your file contains only one image, it will be like this:
NSImageRep *rep = [[image representations] objectAtIndex:0];
NSSize imageSize = NSMakeSize(rep.pixelsWide, rep.pixelsHigh);
where image is your NSImage and imageSize is your image size in pixels.
NSImage size method returns size information that is screen resolution dependent. To get the size represented in the actual file image you need to use an NSImageRep.
Refer nsimage-size-not-real-size-with-some-pictures link and get helped
the direct API gives also the correct results
CGImageRef cgImage = [oldImage CGImageForProposedRect:nil context:context hints:nil];
size_t width = CGImageGetWidth(cgImage);
size_t height = CGImageGetHeight(cgImage);
Apple uses a point system based on DPI to map points to physical device pixels. It doesnt matter what the EXIF says, it matters how many logical screen points your canvas has to display the image.
iOS and OSX perform this mapping for you. The only size you should be concerned about is the size returned from UIImage.size
You cant (read shouldnt have to shouldnt care) do the mapping to device pixels yourself, thats why apple does it.
SWIFT 4
You have to make a NSBitmapImageRep representation of the NSImage to get the correct pixel height and width.
First a this extension to gather a CGImage from the NSImage:
extension NSImage {
#objc var CGImage: CGImage? {
get {
guard let imageData = self.tiffRepresentation else { return nil }
guard let sourceData = CGImageSourceCreateWithData(imageData as CFData, nil) else { return nil }
return CGImageSourceCreateImageAtIndex(sourceData, 0, nil)
}
}
}
Then when you want to get the height and width:
let rep = NSBitmapImageRep(cgImage: (NSImage(named: "Your Image Name")?.CGImage)!)
let imageHeight = rep.size.height
let imageWidth = rep.size.width
i make a extension like this:
extension NSImage{
var pixelSize: NSSize?{
if let rep = self.representations.first{
let size = NSSize(width: rep.pixelsWide, height: rep.pixelsHigh)
return size
}
return nil
}
}

Cocoa Touch - Adding texture with overlay view

I have a set of tiles as UIViews that have a programmable background color, and each one
can be a different color. I want to add texture, like a side-lit bevel, to each one. Can this be done with an overlay view or by some other method?
I'm looking for suggestions that don't require a custom image file for each case.
This may help someone, although this was pieced together from other topics on SO.
To create a beveled tile image with an arbitrary color for normal and for retina display, I made a beveled image in photoshop and set the saturation to zero, making a grayscale image called tileBevel.png
I also created one for the retina display (tileBevel#2x.png)
Here is the code:
+ (UIImage*) createTileWithColor:(UIColor*)tileColor {
int pixelsHigh = 44;
int pixelsWide = 46;
UIImage *bottomImage;
if([UIScreen respondsToSelector:#selector(scale)] && [[UIScreen mainScreen] scale] == 2.0) {
pixelsHigh *= 2;
pixelsWide *= 2;
bottomImage = [UIImage imageNamed:#"tileBevel#2x.png"];
}
else {
bottomImage = [UIImage imageNamed:#"tileBevel.png"];
}
CGImageRef theCGImage = NULL;
CGContextRef tileBitmapContext = NULL;
CGRect rectangle = CGRectMake(0,0,pixelsWide,pixelsHigh);
UIGraphicsBeginImageContext(rectangle.size);
[bottomImage drawInRect:rectangle];
tileBitmapContext = UIGraphicsGetCurrentContext();
CGContextSetBlendMode(tileBitmapContext, kCGBlendModeOverlay);
CGContextSetFillColorWithColor(tileBitmapContext, tileColor.CGColor);
CGContextFillRect(tileBitmapContext, rectangle);
theCGImage=CGBitmapContextCreateImage(tileBitmapContext);
UIGraphicsEndImageContext();
return [UIImage imageWithCGImage:theCGImage];
}
This checks to see if the retina display is used, sizes the rectangle to draw in, picks the appropriate grayscale base image, set the blending mode to overlay, then draws a rectangle on top of the bottom image. All of this is done inside a graphics context bracketed by the BeginImageContext and EndImageContext calls. These set the current context needed by the UIImage drawRect: method. The Core Graphics functions need the context as a parameter, which is obtained by a call to get the current context.
And the result looks like this:
If you want to preserve the alpha channel of the source image, just add this to jim's code before the fill rect:
// Apply mask
CGContextTranslateCTM(tileBitmapContext, 0, rectangle.size.height);
CGContextScaleCTM(tileBitmapContext, 1.0f, -1.0f);
CGContextClipToMask(tileBitmapContext, rectangle, bottomImage.CGImage);
Swift 3 solution, essentially based on Jim's answer with Scriptease's addition, and some minor changes:
class func image(bottomImage: UIImage, topImage: UIImage, tileColor: UIColor) -> UIImage? {
let pixelsHigh: CGFloat = bottomImage.size.height
let pixelsWide: CGFloat = bottomImage.size.width
let rectangle = CGRect.init(x: 0, y: 0, width: pixelsWide, height: pixelsHigh)
UIGraphicsBeginImageContext(rectangle.size);
bottomImage.draw(in: rectangle)
if let tileBitmapContext = UIGraphicsGetCurrentContext() {
tileBitmapContext.setBlendMode(.overlay)
tileBitmapContext.setFillColor(tileColor.cgColor)
tileBitmapContext.scaleBy(x: 1.0, y: -1.0)
tileBitmapContext.clip(to: rectangle, mask: bottomImage.cgImage!)
tileBitmapContext.fill(rectangle)
let theCGImage = tileBitmapContext.makeImage()
UIGraphicsEndImageContext();
if let theImage = theCGImage {
return UIImage.init(cgImage: theImage)
}
}
return nil
}

UIButton image rotation issue

I have a [UIButton buttonWithType:UIButtonTypeCustom] that has an image (or a background image - same problem) created by [UIImage imageWithContentsOfFile:] pointing to a JPG file taken by the camera and saved in the documents folder by the application.
If I define the image for UIControlStateNormal only, then when I touch the button the image gets darker as expected, but it also rotates either 90 degrees or 180 degrees. When I remove my finger it returns to normal.
This does not happen if I use the same image for UIControlStateHighlighted, but then I lose the touch indication (darker image).
This only happens with an image read from a file. It does not happen with [UIImage ImageNamed:].
I tried saving the file in PNG format rather than as JPG. In this case the image shows up in the wrong orientation to begin with, and is not rotated again when touched. This is not a good solution anyhow because the PNG is far too large and slow to handle.
Is this a bug or am I doing something wrong?
I was not able to find a proper solution to this and I needed a quick workaround. Below is a function which, given a UIImage, returns a new image which is darkened with a dark alpha fill. The context fill commands could be replaced with other draw or fill routines to provide different types of darkening.
This is un-optimized and was made with minimal knowledge of the graphics api.
You can use this function to set the UIControlStateHighlighted state image so that at least it will be darker.
+ (UIImage *)darkenedImageWithImage:(UIImage *)sourceImage
{
UIImage * darkenedImage = nil;
if (sourceImage)
{
// drawing prep
CGImageRef source = sourceImage.CGImage;
CGRect drawRect = CGRectMake(0.f,
0.f,
sourceImage.size.width,
sourceImage.size.height);
CGContextRef context = CGBitmapContextCreate(NULL,
drawRect.size.width,
drawRect.size.height,
CGImageGetBitsPerComponent(source),
CGImageGetBytesPerRow(source),
CGImageGetColorSpace(source),
CGImageGetBitmapInfo(source)
);
// draw given image and then darken fill it
CGContextDrawImage(context, drawRect, source);
CGContextSetBlendMode(context, kCGBlendModeOverlay);
CGContextSetRGBFillColor(context, 0.f, 0.f, 0.f, 0.5f);
CGContextFillRect(context, drawRect);
// get context result
CGImageRef darkened = CGBitmapContextCreateImage(context);
CGContextRelease(context);
// convert to UIImage and preserve original orientation
darkenedImage = [UIImage imageWithCGImage:darkened
scale:1.f
orientation:sourceImage.imageOrientation];
CGImageRelease(darkened);
}
return darkenedImage;
}
To fix this you need additional normalization function like this:
public extension UIImage {
func normalizedImage() -> UIImage! {
if self.imageOrientation == .Up {
return self
}
UIGraphicsBeginImageContextWithOptions(self.size, false, self.scale)
self.drawInRect(CGRectMake(0, 0, self.size.width, self.size.height))
let normalized = UIGraphicsGetImageFromCurrentImageContext()
UIGraphicsEndImageContext()
return normalized
}
}
then you can use it like that:
self.photoButton.sd_setImageWithURL(avatarURL,
forState: .Normal,
placeholderImage: UIImage(named: "user_avatar_placeholder")) {
[weak self] (image, error, cacheType, url) in
guard let strongSelf = self else {
return
}
strongSelf.photoButton.setImage(image.normalizedImage(), forState: .Normal
}

AVFoundation crop captured still image according to the preview aspect ratio

My question is mostly similar to this one:
Cropping image captured by AVCaptureSession
I have an application which uses AVFoundation for capturing still images. My AVCaptureVideoPreviewLayer has AVLayerVideoGravityResizeAspectFill video gravity thus making preview picture which is shown to the user to be cropped from the top and from the bottom parts.
When user is pressing "Capture" button, the image actually captured is differs from the preview picture shown to user. My question is how to crop captured image accordingly?
Thanks in advance.
I used UIImage+Resize category provided in here with some new methods I wrote to do the job. I reformatted some code to look better and not tested, but it should work. :))
- (UIImage*)cropAndResizeAspectFillWithSize:(CGSize)targetSize
interpolationQuality:(CGInterpolationQuality)quality {
UIImage *outputImage = nil;
UIImage *imageForProcessing = self;
// crop center square (AspectFill)
if (self.size.width != self.size.height) {
CGFloat shorterLength = 0;
CGPoint origin = CGPointZero;
if (self.size.width > self.size.height) {
origin.x = (self.size.width - self.size.height)/2;
shorterLength = self.size.height;
}
else {
origin.y = (self.size.height - self.size.width)/2;
shorterLength = self.size.width;
}
imageForProcessing = [imageForProcessing normalizedImage];
imageForProcessing = [imageForProcessing croppedImage:CGRectMake(origin.x, origin.y, shorterLength, shorterLength)];
}
outputImage = [imageForProcessing resizedImage:targetSize interpolationQuality:quality];
return outputImage;
}
// fix image orientation, which may wrongly rotate the output.
- (UIImage *)normalizedImage {
if (self.imageOrientation == UIImageOrientationUp) return self;
UIGraphicsBeginImageContextWithOptions(self.size, NO, self.scale);
[self drawInRect:(CGRect){0, 0, self.size}];
UIImage *normalizedImage = UIGraphicsGetImageFromCurrentImageContext();
UIGraphicsEndImageContext();
return normalizedImage;
}