- 在 Android 开发中, Bitmap 是个吃内存大户,稍微操作不当就会 OOM 。虽然现在第三方的图片加载库已经很多,很完善,但是作为一个 Androider 还得知道如何自己进行操作来加载大图。
- 为什么加载图片会很容易造成 OOM 呢,主要是从图片加载到内存说起,假如一个图片的分辨率是 1000*20000,那么这张图片加载的内存中的大致大小为 1000*20000*4 = 80000000 字节,那么就是占用内存为 77 M 左右,这样的话,很容易造成 OOM 。
- 为了不 OOM ,Android 提供了 BitmapFactory.Options 的 inJustDecodeBounds 和 inSimpleSize ,合理使用这些变量可以轻松的加载图片
inJustDecodeBounds
- 通过把该变量设置为 true ,可以在不加载图片的情况下,拿到图片的分辨率。这时,decodeResource 方法返回的 Bitmap 是 null
- 代码
BitmapFactory.Options options = new Options();
options.inJustDecodeBounds = true;
BitmapFactory.decodeResource(getResources(), R.drawable.bigpicture, options);
int outHeight = options.outHeight;
int outWidth = options.outWidth;
inSimpleSize
- 通过 inJustDecodeBounds 拿到了图片的分辨率,那么通过 ImageView 的宽和高与图片的宽高进行比较,只有当图片的宽和高任意一个都比图片的宽和高都小的时候,计算结束。inSimpleSize 的大小,默认值为 1 ,然后按照 2 的倍数增加,假如 inSimpleSize 为 2,那么图片的宽和高就按照1/2缩放,这样加载到内存就缩小了4倍。
- 代码
BitmapFactory.Options options = new Options();
options.inJustDecodeBounds = true;
BitmapFactory.decodeResource(getResources(), R.drawable.bigpicture, options);
int outHeight = options.outHeight;
int outWidth = options.outWidth;
int inSampleSize = 1;
int height = view.getMeasuredHeight();
int width = view.getMeasuredWidth();
if (outHeight > height || outWidth > width) {
int halfHeight = outHeight / 2;
int halfWidth = outWidth / 2;
while ((halfHeight / inSampleSize) >= height || (halfWidth / inSampleSize) >= width) {
inSampleSize *= 2;
}
}
加载大图
private void loadImage(ImageView view) {
BitmapFactory.Options options = new Options();
options.inJustDecodeBounds = true;
BitmapFactory.decodeResource(getResources(), R.drawable.bigpicture, options);
int outHeight = options.outHeight;
int outWidth = options.outWidth;
int inSampleSize = 1;
int height = view.getMeasuredHeight();
int width = view.getMeasuredWidth();
if (outHeight > height || outWidth > width) {
int halfHeight = outHeight / 2;
int halfWidth = outWidth / 2;
while ((halfHeight / inSampleSize) >= height || (halfWidth / inSampleSize) >= width) {
inSampleSize *= 2;
}
}
options.inSampleSize = inSampleSize;
options.inJustDecodeBounds = false;
Bitmap bitmap = BitmapFactory.decodeResource(getResources(), R.drawable.bigpicture, options);
view.setImageBitmap(bitmap);
}
展示巨图局部
- 加载巨图,针对上面方式的话,就无法看清楚了。比如清明上河图,这时候就需要使用到另外的一个类 BitmapRegionDecoder ,通过该类可以展示圈定的矩形区域,这样就可以更加清晰的看到局部了。
- 代码
InputStream inputStream = getResources().getAssets().open("bigpicture.jpg");
BitmapRegionDecoder regionDecoder = BitmapRegionDecoder.newInstance(inputStream, false);
int measuredHeight = view.getMeasuredHeight();
int measuredWidth = view.getMeasuredWidth();
Rect rect = new Rect(0, 0, measuredWidth, measuredHeight);
Bitmap bitmap = regionDecoder.decodeRegion(rect, new Options());
view.setImageBitmap(bitmap);
- 以上就可以在 ImageView 中展示图片的局部。
滑动显示巨图
- 上面功能实现了展示巨图的局部,但是想要通过滑动显示巨图的其他区域,就需要自定义 View , 重写 onTouchEvent() 方法,根据手指滑动的距离,重新计算 Rect 的区域,来实现加载大图布局。
- 几个要点:
- 在 onMeasure 中拿到测量后的大小,设置给 Rect
- 在 onTouchEvent() 方法中,计算滑动的距离,然后设置给 Rect
- 设置了新的显示区域以后,调用 invalidate() 方法, 请求绘制,这时候会调用 onDraw() 方法
- 在 onDraw() 方法中根据 Rect 拿到需要显示的局部 Bitmap, 通过 Canvas 绘制回来。
public class BigImageView extends View {
private int slideX = 0;
private int slideY = 0;
private BitmapRegionDecoder bitmapRegionDecoder;
private Paint paint;
private int mImageWidth;
private int mImageHeight;
private Options options;
private Rect mRect;
public BigImageView(Context context) {
this(context, null);
}
public BigImageView(Context context, @Nullable AttributeSet attrs) {
this(context, attrs, 0);
}
public BigImageView(Context context, @Nullable AttributeSet attrs, int defStyleAttr) {
super(context, attrs, defStyleAttr);
init();
}
private void init() {
mRect = new Rect();
paint = new Paint();
try {
InputStream inputStream = getResources().getAssets().open("bigpicture.jpg");
options = new Options();
options.inJustDecodeBounds = true;
BitmapFactory.decodeStream(inputStream, null, options);
mImageWidth = options.outWidth;
mImageHeight = options.outHeight;
options.inJustDecodeBounds = false;
options.inPreferredConfig = Config.RGB_565;
bitmapRegionDecoder = BitmapRegionDecoder.newInstance(inputStream, false);
} catch (IOException e) {
e.printStackTrace();
}
}
float downX = 0;
float downY = 0;
@Override
public boolean onTouchEvent(MotionEvent event) {
switch (event.getAction()) {
case MotionEvent.ACTION_DOWN:
downX = event.getRawX();
downY = event.getRawY();
break;
case MotionEvent.ACTION_MOVE:
float moveX = event.getRawX();
float moveY = event.getRawY();
slideX = (int) (moveX - downX);
slideY = (int) (moveY - downY);
if (mImageWidth > getWidth()) {
mRect.offset(-slideX, 0);
if (mRect.right > mImageWidth) {
mRect.right = mImageWidth;
mRect.left = mRect.right - getWidth();
}
if (mRect.left < 0) {
mRect.left = 0;
mRect.right = getWidth();
}
invalidate();
}
if (mImageHeight > getHeight()) {
mRect.offset(0, -slideY);
if (mRect.bottom > mImageHeight) {
mRect.bottom = mImageHeight;
mRect.top = mRect.bottom - getHeight();
}
if (mRect.top < 0) {
mRect.top = 0;
mRect.bottom = getHeight();
}
invalidate();
}
downX = moveX;
downY = moveY;
break;
default:
break;
}
return true;
}
@Override
protected void onMeasure(int widthMeasureSpec, int heightMeasureSpec) {
super.onMeasure(widthMeasureSpec, heightMeasureSpec);
mRect.left = 0;
mRect.right = getMeasuredWidth() + mRect.left;
mRect.top = 0;
mRect.bottom = getMeasuredHeight() + mRect.top;
}
@Override
protected void onDraw(Canvas canvas) {
super.onDraw(canvas);
Bitmap bitmap = bitmapRegionDecoder.decodeRegion(mRect, options);
canvas.drawBitmap(bitmap, 0, 0, paint);
}
}