13

目标：取一个 UIImage，在中间裁剪一个正方形，将正方形的大小更改为 320x320 像素，将图像切成 16 个 80x80 的图像，将 16 个图像保存在一个数组中。

这是我的代码：

CGImageRef originalImage, resizedImage, finalImage, tmp;
float imgWidth, imgHeight, diff;
UIImage *squareImage, *playImage;
NSMutableArray *tileImgArray;
int r, c;

originalImage = [image CGImage];

imgWidth = image.size.width;
imgHeight = image.size.height;
diff = fabs(imgWidth - imgHeight);

if(imgWidth > imgHeight){
    resizedImage = CGImageCreateWithImageInRect(originalImage, CGRectMake(floor(diff/2), 0, imgHeight, imgHeight));
}else{
    resizedImage = CGImageCreateWithImageInRect(originalImage, CGRectMake(0, floor(diff/2), imgWidth, imgWidth));
}
CGImageRelease(originalImage);

squareImage = [UIImage imageWithCGImage:resizedImage];      
if(squareImage.size.width != squareImage.size.height){
    NSLog(@"image cutout error!");
    //*code to return to main menu of app, irrelevant here
}else{
    float newDim = squareImage.size.width;
    if(newDim != 320.0){
        CGSize finalSize = CGSizeMake(320.0, 320.0);
        UIGraphicsBeginImageContext(finalSize);
        [squareImage drawInRect:CGRectMake(0, 0, finalSize.width, finalSize.height)];
        playImage = UIGraphicsGetImageFromCurrentImageContext();
        UIGraphicsEndImageContext();
    }else{
        playImage = squareImage;
    }
}

finalImage = [playImage CGImage];
tileImgArray = [NSMutableArray arrayWithCapacity:0];
for(int i = 0; i < 16; i++){
    r = i/4;
    c = i%4;
    //*
    tmp = CGImageCreateWithImageInRect(finalImage, CGRectMake(c*tileSize, r*tileSize, tileSize, tileSize));
    [tileImgArray addObject:[UIImage imageWithCGImage:tmp]];
}

当原始（可变图像）的较小尺寸大于或小于 320 像素时，代码可以正常工作。当它正好是 320 时，生成的 80x80 图像几乎完全是黑色的，有些边缘有几个像素（我无法确定）来自原始图像。

我通过直接显示完整图像进行了测试：

[UIImage imageWithCGImage:finalImage];

间接地：

[UIImage imageWithCGImage:CGImageCreateWithImageInRect(finalImage, CGRectMake(0, 0, 320, 320))];

在这两种情况下，显示都有效。只有当我尝试切出图像的某些部分时才会出现问题。

4

2 回答 2

5

经过更多的实验，我找到了以下解决方案（虽然我仍然不知道为什么它没有像最初写的那样工作。）但是无论如何，即使不需要调整大小，在调整大小代码到位后切片仍然有效：

if(newDim != 320.0){
            CGSize finalSize = CGSizeMake(320.0, 320.0);
            UIGraphicsBeginImageContext(finalSize);
            [squareImage drawInRect:CGRectMake(0, 0, finalSize.width, finalSize.height)];
            playImage = UIGraphicsGetImageFromCurrentImageContext();
            UIGraphicsEndImageContext();
}else{
            CGSize finalSize = CGSizeMake(320.0, 320.0);
            UIGraphicsBeginImageContext(finalSize);
            [squareImage drawInRect:CGRectMake(0, 0, finalSize.width, finalSize.height)];
            playImage = UIGraphicsGetImageFromCurrentImageContext();
            UIGraphicsEndImageContext();
}

任何人都知道为什么会这样？

PS 是的，这里不再需要 if/else。但是，在我知道它会起作用之前将其删除是愚蠢的。

于 2008-10-30T03:11:14.540 回答
1

只是出于好奇，当您知道要在其中放入 16 个东西时，为什么要使可变数组的边界为 0？

好吧，除此之外，我已经尝试了您用于调整大小和切片的基本技术（我不需要裁剪，因为我正在处理已经是方形的图像）并且我无法在模拟器。您可能想尝试将代码分解为三个单独的函数（裁剪为正方形、调整大小和切片），然后分别测试这三个函数，以便找出三个步骤中的哪一个导致问题（即输入图像您已经在普通的图形程序中进行操作，而不是使用目标 c，然后检查您得到的结果！）。

我将在下面附上我的调整大小和切片功能版本，希望对您有所帮助。很高兴看到你的版本，因为我不必一次就自己找到所有的方法。:)

需要注意的是，提到的二维数组是我自己用 NSMutableArrays 构建的类，但您可以轻松实现自己的版本或使用平面 NSMutableArray。;)

// cut the given image into a grid of equally sized smaller images
// this assumes that the image can be equally divided in the requested increments
// the images will be stored in the return array in [row][column] order
+ (TwoDimensionalArray *) chopImageIntoGrid : (UIImage *) originalImage : (int) numberOfRows : (int) numberOfColumns
{   
// figure out the size of our tiles
int tileWidth = originalImage.size.width / numberOfColumns;
int tileHeight = originalImage.size.height / numberOfRows;

// create our return array
TwoDimensionalArray * toReturn = [[TwoDimensionalArray alloc] initWithBounds : numberOfRows 
                                                                             : numberOfColumns];

// get a CGI image version of our image
CGImageRef cgVersionOfOriginal = [originalImage CGImage];

// loop to chop up each row
for(int row = 0; row < numberOfRows ; row++){
    // loop to chop up each individual piece by column
    for (int column = 0; column < numberOfColumns; column++)
    {
        CGImageRef tempImage = 
                CGImageCreateWithImageInRect(cgVersionOfOriginal, 
                                             CGRectMake(column * tileWidth, 
                                                        row * tileHeight, 
                                                        tileWidth, 
                                                        tileHeight));
        [toReturn setObjectAt : row : column : [UIImage imageWithCGImage:tempImage]];
    }
}

// now return the set of images we created
return [toReturn autorelease];
}

// this method resizes an image to the requested dimentions
// be a bit careful when using this method, since the resize will not respect
// the proportions of the image
+ (UIImage *) resize : (UIImage *) originalImage : (int) newWidth : (int) newHeight
{   
// translate the image to the new size
CGSize newSize = CGSizeMake(newWidth, newHeight); // the new size we want the image to be
UIGraphicsBeginImageContext(newSize); // downside: this can't go on a background thread, I'm told
[originalImage drawInRect : CGRectMake(0, 0, newSize.width, newSize.height)];
UIImage* newImage = UIGraphicsGetImageFromCurrentImageContext(); // get our new image
UIGraphicsEndImageContext();

// return our brand new image
return newImage;
}

伊娃·希弗

于 2009-06-07T18:14:49.743 回答