It is worth pointing out that the CLBP features are extracted based on uniform and rotation invariant pattern mapping, while the LBP features without pattern mapping are used as a baseline to be compared with.
Intuitively, LBP can be understood as only sign pattern component of CLBP operators.
In order to achieve a fair comparison to CLBP and LBP, only second-order tetra patterns [LTrP.sup.2]([g.sub.c]) based on the same configurations of R = 1 and N = 8 are employed in this study and then converted the binary coding of tetra and magnitude pattern (for example, see the bottom of Figure 3) to decimal numeral.
Two global features (i.e., Haralick texture and DNA distribution features) and three local pattern features (i.e., LBP, CLBP, and LTrP features) were employed to describe IHC images in this study.
We use CLBP to obtain three kinds of codes, which are CLBP_S, CLBP_C and CLBP_M, for each pixel.
These compared methods are DRLBP , CLBP , LDBP , PLBP , PRICoLBP , MDLBP  LTrP  and DFD .
We involve LBP and CLBP in our feature extraction step.
It is notable that Dual-CLBP + CLBP performs worse than others on Set3 and Set4.
2.3 A completed model of the local binary pattern (CLBP)
The introduction of CLBP_M and CLBP C increases the distinguishable information of texture.
First, the CLBP_S, CLBP_M, and CLBP_C of the original image are calculated to obtain the mapping of CLBP, and then the statistical histogram is calculated.
Therefore, we can use slide windows with different sizes and calculate their CLBP histogram to get different coarse-grained information.