The planar near-field scanning frame requires high flatness, which requires error measurement and compensation after assembly. This paper systematically analyzes the commonly used parameter identification methods for planar near-field, and determines the use of laser tracker as the error identification method. Secondly, according to the error transfer matrix derived from the measured data, the six-degree-of-freedom errors of two coordinate systems are obtained. Finally, based on the measured data, the error compensation of scanning location and flatness compensation of scanning plane are respectively carried out, and the compensation effect of weighted least square method and unweighted least square method is compared. By analyzing the flatness of the data, the XY plane of the scanning frame is adjusted by means of hardware, software, error source and overall error compensation to meet the higher plane requirements.