An image processing method, comprising the following steps: obtaining a plurality of first luminance values, wherein the plurality of first luminance values corresponds to a first subpixel group comprising a target subpixel and a plurality of adjacent subpixels; and performing a subpixel rendering conversion on a target luminance value of the plurality of first luminance values corresponding to the target subpixel according to a weighting matrix and all of the plurality of first luminance values, so that the target luminance value is converted to a rendered luminance value, wherein the weighting matrix comprises a plurality of weighting parameters corresponding to the first subpixel group, and the weighting matrix is time-variant.