/*********************************
函數(shù)聲明：初始化偏置圖像
參數(shù)：NONE
注釋：NONE
測試：NONE
**********************************/
void PatchMatch::InitOff(Mat Mask, Mat &Off)
{
	//為方便起見，將所有的都附上，要求不能賦值到非搜索區(qū)域
	//初始化格式
	Off = Mat(Mask.size(), CV_32FC2, Scalar::all(0));//2維無符號32位精度浮點數(shù)

	for (int i = 0; i < Mask.rows; i++)
	{
		for (int j = 0; j < Mask.cols; j++)
		{
			//不考慮search區(qū)域，沒有破損，他們的最佳偏移向量當然是0，自己
			if (Mask.at<uchar>(i, j) == search)
			{
				Off.at<Vec2f>(i, j)[0] = 0;  //<Vec2f> 向量，2維，浮點數(shù)
				Off.at<Vec2f>(i, j)[1] = 0;
			}
			else//處理hole，采用隨機偏置  
			{
				//先初始化2個偏置數(shù)r_col,r_row
				int r_col = rand() % Mask.cols; //rand（）產(chǎn)生隨機數(shù)，主要是產(chǎn)生一個偏置的初始值
				int r_row = rand() % Mask.rows;
				r_col = r_col + j < Mask.cols ? r_col : r_col - Mask.cols;//邊界檢測
				r_row = r_row + i < Mask.rows ? r_row : r_row - Mask.rows;

				//為什么要有這個循環(huán)？因為一次的隨機賦值，很可能會出現(xiàn)偏置后的塊跑到破損區(qū)域，或者是超出限定搜索框的邊界
				while (
					!(Mask.at<uchar>(r_row + i, r_col + j) == search	//這里加上I，j，是因為他是A投影到B中的搜索偏置
						&& abs(r_row) < searchrowratio*Mask.rows))	//searchrowratio=0.5，搜索的時候,確保r_row偏置不會太遠，一定是在原圖像的大小里	
				{
					r_col = rand() % Mask.cols;
					r_row = rand() % Mask.rows;

					//邊界檢測
					r_col = r_col + j < Mask.cols ? r_col : r_col - Mask.cols;
					r_row = r_row + i < Mask.rows ? r_row : r_row - Mask.rows;
				}

				//賦偏置值
				Off.at<Vec2f>(i, j)[0] = r_row;
				Off.at<Vec2f>(i, j)[1] = r_col;
			}
		}
	}
}

//以塊為單位，用所有像素點的相同顏色通道的差平方來簡單判斷相似度
float PatchMatch::Distance(Mat Dst, Mat Src)
{
	float distance = 0;

	for (int i = 0; i < Dst.rows; i++)
	{
		for (int j = 0; j < Dst.cols; j++)
		{
			for (int k = 0; k < 3; k++)//K=3個顏色通道
			{
				int tem = Src.at < Vec3b >(i, j)[k] - Dst.at < Vec3b >(i, j)[k];
				distance += tem * tem;//差平方
			}
		}
	}

	return distance;
}

//迭代第一步：傳播
//（now_row, now_col）：patch里的像素
//odd：當前迭代次
void PatchMatch::Propagation(Mat Dst, Mat Src, Mat Mask, Mat &Off, int row, int col,int odd)
{
	Mat DstPatch = GetPatch(Dst, row, col);//獲取長度為 patchsize = 3 的邊界框, (row, col)代表的是中心像素點坐標

	if (odd % 2 == 0)//偶次迭代
	{
		//提取(row, col)的match塊
		Mat SrcPatch = GetPatch(Src, row + Off.at < Vec2f >(row, col)[0],
			col + Off.at < Vec2f >(row, col)[1]);

		//提取(row, col-1)的match塊
		Mat LSrcPatch = GetPatch(Src, row + Off.at < Vec2f >(row, col - 1)[0],
			col - 1 + Off.at < Vec2f >(row, col - 1)[1]);

		//提取(row-1, col)的match塊
		Mat USrcPatch = GetPatch(Src,
			row - 1 + Off.at < Vec2f >(row - 1, col)[0],
			col + Off.at < Vec2f >(row - 1, col)[1]);

		//返回上面4個塊最相似的塊的代表數(shù)字，用于switch判斷
		int location = GetMinPatch1(DstPatch, SrcPatch, LSrcPatch, USrcPatch);

		//利用上面的信息更新像素點的偏置地圖
		switch (location)
		{
			//若是1則不更新
		case 2:
			Off.at < Vec2f >(row, col)[0] = Off.at < Vec2f >(row, col - 1)[0];
			Off.at < Vec2f >(row, col)[1] = Off.at < Vec2f >(row, col - 1)[1] - 1;
			break;
		case 3:
			Off.at < Vec2f >(row, col)[0] = Off.at < Vec2f >(row - 1, col)[0] - 1;
			Off.at < Vec2f >(row, col)[1] = Off.at < Vec2f >(row - 1, col)[1];
			break;
		}
	}

	else//奇數(shù)次迭代
	{
		Mat SrcPatch = GetPatch(Src, row + Off.at < Vec2f >(row, col)[0],
			col + Off.at < Vec2f >(row, col)[1]);
		Mat RSrcPatch = GetPatch(Src, row + Off.at < Vec2f >(row, col + 1)[0],
			col + 1 + Off.at < Vec2f >(row, col + 1)[1]);
		Mat DSrcPatch = GetPatch(Src,
			row + 1 + Off.at < Vec2f >(row + 1, col)[0],
			col + Off.at < Vec2f >(row + 1, col)[1]);

		int location = GetMinPatch1(DstPatch, SrcPatch, RSrcPatch, DSrcPatch);
		switch (location)
		{
		case 2:
			Off.at < Vec2f >(row, col)[0] = Off.at < Vec2f >(row, col + 1)[0];
			Off.at < Vec2f >(row, col)[1] = Off.at < Vec2f
			>(row, col + 1)[1] + 1;
			break;
		case 3:
			Off.at < Vec2f >(row, col)[0] = Off.at < Vec2f
			>(row + 1, col)[0] + 1;
			Off.at < Vec2f >(row, col)[1] = Off.at < Vec2f >(row + 1, col)[1];
			break;
		}
	}
}

//迭代第二步：隨機搜索
//（row,col）=（now_row, now_col）：修復patch里的像素
void PatchMatch::RandomSearch(Mat Dst, Mat Src, Mat Mask, Mat &Off, int row, int col)
{
	Mat DstPatch = GetPatch(Dst, row, col);//獲取修復基準框，在框內(nèi)操作

	//迭代指數(shù)
	int attenuate = 0;

	while (true)
	{
		//獲取隨機參數(shù)，在 [-1;1] 間
		float divcol = rand() % 2000 / 1000.0f - 1.0f;
		float divrow = rand() % 2000 / 1000.0f - 1.0f;

		//減小框大小的公式，𝑢_𝑖=𝑣_0+𝑤*𝛼^𝑖*𝑅_𝑖
		//行列分別處理，MaxWindow：原始框?qū)挾?；divcol：隨機系數(shù)；pow(A,B):A的B次方。隨迭代次數(shù)而變小的縮小系數(shù)；RandomAttenuation=0.5；
		float veccol = MaxWindow * pow(RandomAttenuation, attenuate)* divcol;
		float vecrow = MaxWindow * pow(RandomAttenuation, attenuate)* divrow;

		float length = sqrt(veccol * veccol + vecrow * vecrow);
		//如果低于1個像素，沒有意義，直接結(jié)束整個循環(huán)，對下一個像素處理
		if (length < 1)
			break;

		//x方向,前2項指向(row, col)的match塊，后面是公式的后一項
		int nowrow = row + Off.at < Vec2f >(row, col)[0] + vecrow;
		//y方向
		int nowcol = col + Off.at < Vec2f >(row, col)[1] + veccol;

		//判斷隨機搜索的patch不越界，在search內(nèi)
		if (nowcol >= 0 && nowcol <= Off.cols - 1 && nowrow >= 0
			&& nowrow <= Off.rows - 1
			&& Mask.at < uchar >(nowrow, nowcol) == search
			&& abs(nowrow - row) < searchrowratio * Mask.rows)//abs：絕對值
		{
			//取出原來的match塊
			Mat SrcPatch1 = GetPatch(Src, Off.at < Vec2f >(row, col)[0] + row,
				Off.at < Vec2f >(row, col)[1] + col);
			//取出現(xiàn)在的隨機match塊
			Mat SrcPatch2 = GetPatch(Src, nowrow, nowcol);

			//對比相似性，找出最好的塊
			int location = GetMinPatch2(DstPatch, SrcPatch1, SrcPatch2);

			//結(jié)合最好的相似塊給像素新的偏置值
			switch (location)
			{
			case 2:
				Off.at < Vec2f >(row, col)[1] = nowcol - col;
				Off.at < Vec2f >(row, col)[0] = nowrow - row;
				break;
			}
		}

		//迭代指數(shù)增加
		attenuate++;
	}
}