Authors: ABC Fintech

Description: We evaluate the fontsize distribution of the input image. And then detect the text lines using pixellink of Multiple scales.