Terchertz (THz) transmission imaging is characterized by its harmless nature to human body and high penetration capability for non-metallic materials, making it suitable for applications in security inspection and biomedical fields. The key to improve the measurement accuracy of THz line scan camera lies in the accuracy of camera parameters. This paper proposes an improved Dingo optimization algorithm (DOA) to optimize the camera parameters obtained by the Draréni calibration method. The improved DOA enhances the global search capability by modifying the feeding and survival behaviors in the traditional DOA, and introducing the individual updating strategy of particle swarm optimization algorithm and grey wolf algorithm. A THz transmission scanning imaging device is built for imaging, and 10 images from different positions are selected as test objects. The camera's initial intrinsic and extrinsic parameters are obtained using the Draréni calibration method. Finally, the particle swarm optimization algorithm, DOA, and the improved DOA proposed in this paper are used respectively for camera internal parameter optimization. The experimental results show that the proposed algorithm in this paper reduces the average reprojection error by 33.41%, 21.35%, 12.62% respectively compared with the traditional Draréni calibration method, the particle swarm optimization algorithm, and the standard DOA, which demonstrates that the proposed algorithm is stable and can significantly improve the accuracy of camera calibration.