This paper discusses a kind of statistical feature based image interpolation method at first.Using the covariance matrix and covariance vector in low-resolution image

one can obtain the interpolation coefficients

which can adapt edge orientation.In order to employ statistical feature base interpolation in the area of color image interpolation

this paper present the following methods to increase the calculating speed:estimating the interpolation coefficients only based on Y image

and applying them to R

G

and B images interpolation simultaneously;For edge pixels

statistical feature based image interpolation method is exploited

while for non-edge pixels

we only use simple bilinear interpolation.This method can be called mixture interpolation.Benefit from these measures

one can get high calculating speed while keeping the image quality unchanged.The experiments show that this method is both effective for calculating speed and interpolated image quality.