Image interpolation apparatus and method

Abstract

The present invention is an interpolation device and method. According to the present invention, a plurality of pixels close to a pixel to be interpolated are sequentially set to a central pixel. When a pixel difference closest to a threshold value among pixel differences between the central pixel and a plurality of pixels around the central pixel belongs to a quasi-edge decision range, an edge interpolation and a bilinear interpolation are mixed to interpolate the pixel to be interpolated. As a result, it is possible to prevent reduction in image quality due to an edge verdict caused by unstable input image signal and perform a stable interpolation operation.

1 . An interpolation device, comprising:
 an edge detector sequentially setting four pixels close to a pixel to be interpolated and calculating pixel differences between a central pixel and eight pixels around the central pixel to detect a temporary edge between the interpolation pixel and the close pixels;   an edge direction modifier for transforming a temporary edge detected from the edge detector to generate a predetermined transformed edge information;   an interpolation mix ratio calculator for calculating an interpolation ratio according to a relation of the pixel differences, and the minimum threshold value and the maximum threshold value;   a non-linear interpolation coefficient generator for generating a non-linear interpolation coefficient from edge information from the edge direction modifier and the interpolation mix ratio calculator and outputting an edge detection signal showing whether an edge is between the interpolation pixel and the close pixels;   a linear interpolator for linearly interpolating the interpolation pixel;   a non-linear interpolator for non-linearly interpolating the interpolation pixel referring to the non-linear interpolation coefficient; and   a data selector for outputting any one pixel value interpolated in the linear interpolator and the non-linear interpolator as a pixel value of the pixel to be interpolated in response to the edge detection signal.   
   
   
       2 . The interpolation device of  claim 1 , wherein when the pixel difference closest to the minimum threshold value among the pixel differences is larger than the minimum lowest limit value and smaller than the minimum upper value, the edge interpolation ratio is
   (TH_L+LU−dm)*m   where, TH_L represents the minimum threshold value, LU represents the minimum upper limit value, dm represents a pixel difference closest to the minimum threshold value, and m represents a roll-off of the upper limit value and the lowest limit value.   
   
   
       3 . The interpolation device of  claim 1 , wherein when the pixel difference closest to the maximum threshold value among the pixel differences is larger than the maximum lowest limit value and smaller than the maximum upper value, the edge interpolation ratio is
   (TH_H+HL−dm)*m   where, TH_H represents the maximum threshold value, HL represents the lowest limit value, dm represents a pixel difference closest to the maximum threshold value, and m represents a roll-off of the upper limit value and the lowest limit value.   
   
   
       4 . The interpolation device of  claim 1 , wherein the non-linear interpolator outputs PNW*(W_NW)+PNE*(W_NE)+PSW*(W_SW)+PSE*(W_SE) to an interpolation value of the pixel to be interpolated
 where, PNW, PNE, PSW and PSE are four pixels close to the pixel to be interpolated, and W_NW, W_NE, W_SW and W_SE are interpolation coefficients, which are calculated in the step of calculating the interpolation coefficient.   
   
   
       5 . The interpolation device of  claim 4 , wherein the interpolation coefficients W_NW, W_NE, W_SW and W_SE are
     W   —   NW=t ( WNWE )+(1− t ) WNWB          W   —   NE=t ( WNEE )+(1− t ) WNEB          W   —   SW=t ( WSWE )+(1− t ) WSWB          W   —   SE=t ( WSEE )+(1− t ) WSEB      where, t represents a mixing ratio calculated in the step of calculating the mixing ratio, WNWE, WNEE, WSWE and WSEE are edge interpolation coefficients calculated from a position of a coordinate transformation of the pixel to be interpolated, WNWB, WNEB, WSWB and WSEB are bilinear interpolation coefficients, and the bilinear interpolation coefficients are
     WNWB =(1− KX )*(1− KY ) 
     WNEB=KX *(1− KY ) 
     WSWB =(1− KX )* KY    
     WSEB=KX*KY    
   where, KX and KY are respectively a position of a pixel to be interpolated.   
   
   
       6 . The interpolation device of  claim 5 , wherein the mixing ratio (t) is 1 in the case that a pixel difference closest to the minimum threshold value among the pixel differences is the same as or smaller than the minimum lowest limit value, and
 wherein the mixing ratio (t) is 0 in the case that a pixel difference closest to the minimum threshold value among the pixel differences is smaller than the maximum lowest limit value.   
   
   
       7 . The interpolation device of  claim 6 , wherein in the case that a pixel difference closest to the maximum threshold value among the pixel differences is larger than or the same as the maximum upper limit value, the mixing ratio (t) is 1, and
 wherein in case that a pixel difference closest to the maximum threshold value among the pixel differences is larger than the maximum upper limit value, the mixing ratio (t) is 0.