In order to obtain a wide field of view(FOV) and high spatial resolution hyperspectral image while the scanning mirror swinging, the channel separation, FOV and mechanical splicing are required for the geostationary full spectrum hyperspectral imager. If the sub-FOV images are directly spliced, the spectral information of the feature will be distorted because of the different geometric deformation between the sub-FOVs. It was assumed that the difference of the corresponding image points′ coordinates between the sub-FOVs was caused by the changes of each sub-FOVs′ internal parameters in the paper. A block adjustment method based on the panoramic distortion tangent correction formula was proposed to achieve FOV stitching in the image side coordinate system. The FOV stitching experiment was completed on the basis of four sub-FOVs simulation images in the visible to near-infrared channel, and the stitching accuracy of 0.72 pixels was achieved. The splicing accuracy of the proposed method is close to the block adjustment method′s based on the collinear conditional equation. The algorithm is simpler and suitable for primary image production without the assistance of control points and DEM.