近期也是由于项目的原因，所以有机会接触到这边的算法。 此文重点不是怎样实现的思路和原理， 有须要的同事能够借鉴一下

废话不多说，直接上代码：

public class MyImage extends ImageView {    private DisplayMetrics dm;    private int bodyImageViewHeight = 0;    private int bodyImageViewWidth = 0;    public MyImage(Context context) {        super(context);    }    public MyImage(Context context, AttributeSet attrs) {        super(context, attrs);        dm = context.getResources().getDisplayMetrics();    }    @Override    public boolean onTouchEvent(MotionEvent event) {        int x = (int) event.getX();        int y = (int) event.getY();        //初始化各个部位的范围        initParametersForRegion();        if(isTouchPointInTransparent(x, y)) {            Log.e("mcoy", "the point is in transparent area now");        } else {            positionType position = pointToPosition(x, y);            Log.e("mcoy", "the position is " + position);        }        return super.onTouchEvent(event);    }    private enum positionType{        REGION_FRONT_HAND,        REGION_FRONT_HEAD,        REGION_FRONT_CHEST,        REGION_FRONT_WAIST,        REGION_FRONT_LEG    }    private positionType pointToPosition(int x, int y) {        if(x < mHandX1 || x > mHandX2)            return positionType.REGION_FRONT_HAND;        else if (y < mHeadY)            return positionType.REGION_FRONT_HEAD;        else if(y < mChestY)            return positionType.REGION_FRONT_CHEST;        else if(y < mWaistY)            return positionType.REGION_FRONT_WAIST;        else            return positionType.REGION_FRONT_LEG;    }    /**     *     * @param x     * @param y     * @return 推断点击区域是否在透明区域     */    private boolean isTouchPointInTransparent(int x, int y) {        int paddingLeft = this.getPaddingLeft();        int paddingTop = this.getPaddingTop();        int imageHeight = this.getHeight();        int imageWidth = this.getWidth();        Drawable drawable = this.getDrawable();        Bitmap bitmap = ((BitmapDrawable) drawable).getBitmap();        int intrinsicHeight = drawable.getIntrinsicHeight();        int intrinsicWidth = drawable.getIntrinsicWidth();        int locationInBitmapX = (x - paddingLeft) * intrinsicWidth / imageWidth;        int locationInBitmapY = (y - paddingTop) * intrinsicHeight / imageHeight;        int pixel = bitmap.getPixel(locationInBitmapX, locationInBitmapY);        Log.e("mcoy", "x is " + x);        Log.e("mcoy", "y is " + y);        Log.e("mcoy", "imageHeight is " + imageHeight);        Log.e("mcoy", "imageWidth is " + imageWidth);        Log.e("mcoy", "intrinsicHeight is " + intrinsicHeight);        Log.e("mcoy", "intrinsicWidth is " + intrinsicWidth);        Log.e("mcoy", "locationInBitmapX is " + locationInBitmapX);        Log.e("mcoy", "locationInBitmapY is " + locationInBitmapY);        Log.e("mcoy", "actualBitmapX is " + locationInBitmapX / dm.density);        Log.e("mcoy", "actualBitmapY is " + locationInBitmapY / dm.density);        Log.e("mcoy", "pixel is " + pixel);        return pixel == 0;    }    private int mHeadY;    private int mHandX1;    private int mHandX2;    private int mChestY;    private int mWaistY;    //下面数据须要同UI同事沟通好每一个部位定义的范围。或者自己手动量一下    private final int HEAD_AREA = 130;    private final int LEFT_HAND_AREA = 126;    private final int RIGHT_HAND_AREA = 280;    private final int CHEST_AREA = 260;    private final int WAIST_AREA = 417;    private void initParametersForRegion() {        if(bodyImageViewHeight != this.getHeight()) {            bodyImageViewHeight = this.getHeight();            bodyImageViewWidth = this.getWidth();            int imageIntrinsicHeight = this.getDrawable().getIntrinsicHeight();            int imageIntrinsicWidht = this.getDrawable().getIntrinsicWidth();            Log.e("danny", "bodyImageViewHeight is " + bodyImageViewHeight);            Log.e("danny", "bodyImageViewWidth is " + bodyImageViewWidth);            Log.e("danny", "imageIntrinsicHeight is " + imageIntrinsicHeight);            Log.e("danny", "imageIntrinsicWidht is " + imageIntrinsicWidht);            mHeadY = DensityUtil.dip2px(getContext(), HEAD_AREA) * bodyImageViewHeight / imageIntrinsicHeight                    + this.getPaddingTop();            mHandX1 = DensityUtil.dip2px(getContext(), LEFT_HAND_AREA) * bodyImageViewWidth / imageIntrinsicWidht                    + this.getPaddingLeft();            mHandX2 = DensityUtil.dip2px(getContext(), RIGHT_HAND_AREA) * bodyImageViewWidth / imageIntrinsicWidht                    + this.getPaddingLeft();            mChestY = DensityUtil.dip2px(getContext(), CHEST_AREA) * bodyImageViewHeight / imageIntrinsicHeight                    + this.getPaddingTop();            mWaistY = DensityUtil.dip2px(getContext(), WAIST_AREA) * bodyImageViewHeight / imageIntrinsicHeight                    + this.getPaddingTop();            Log.e("danny", "mHeadY is " + mHeadY);            Log.e("danny", "mHandX1 is " + mHandX1);            Log.e("danny", "mHandX2 is " + mHandX2);            Log.e("danny", "mChestY is " + mChestY);            Log.e("danny", "mWaistY is " + mWaistY);        }    }}

代码非常easy， 就是一个自己实现的ImageView， 可是有几点须要注意的点：

1 图片须要跟UI同事沟通好对应坐标(认真看完代码之后就会知道原因), 假设换图片的话。那HEAD_AREA等值可能须要又一次定义

2 算法思路---在不同的desity的的设备上。 我们都能够算出图片相应的drawable的width/height，而图片缩放或者是放大是依照比例缩放/放大， 因此我们能够得出例如以下等式：

touchX / imageWidth = positonInBitmap / drawable.getIntransicWidth ;

==> positonInBitmap = touchX * drawable.getIntransicWidth / imageWidth;

touchX是点击在ImageView上的x坐标；imageWidth是图片的宽度。positionInBitmap是touchX映射在ImageView所相应drawable上的x坐标(这句话有点拗口，能够细致想一下)

下面是xml文件：

也有一点须要注意：

须要将adjustViewBounds属性设置为true。这样会调整ImageView的界限来保持图像纵横比不变

demo下载链接：http://download.csdn.net/detail/zxm317122667/9000969