UNIT: Unifying Image and Text Recognition in One Vision Encoder