From: Estimation of Bayer CFA pattern configuration based on singular value decomposition
Input: A suspicious image. | |
Input parameters: Square block size M and truncated singular value point t. | |
Output: Bayer CFA pattern configuration, C b | |
1.Choose the image block M located at the center of the given image. | |
2.For each color component, decompose the block into four sub-blocks using (1). | |
3.Compute \( {\boldsymbol{D}}_i^{m_{\boldsymbol{D}}(i)} \) and \( {\boldsymbol{F}}_i^{m_{\boldsymbol{F}}(i)} \) using (2) and (3), respectively. | |
4.Compute \( {S}_{D, i}^{m_D(i)} \) and \( {S}_{F, i}^{m_F(i)} \) using (5) and (6), respectively. | |
5.Compute \( {V}_k^{\boldsymbol{D}} \) and \( {V}_k^{\boldsymbol{F}} \) using (7) and (8), respectively. | |
6.Determine candidate index \( \tilde{b} \) using (9). | |
7.Determine Bayer CFA pattern index b using (10). |