With the continuous development of space remote sensing instruments and the further improvement of the quantitative application requirements of remote sensing products, the necessity and importance of high-precision calibration of space remote sensing instruments are increasingly apparent. The moon's reflectance changes by 10^-8/year, and its high stability in reflectance, as well as the repeatability of radiance at the same observation geometry angle, make the moon very suitable as a reference radiation source and a hot research topic in the field of radiation calibration at home and abroad. However, the key to achieving this function is the need for a large amount of ground-based observation data to support the establishment of a high-precision lunar radiation model. Therefore, it is urgent to develop radiation instruments that have long-term precise tracking and measurement capabilities on the moon. Accurate test data and laboratory radiation calibration are two key factors that directly affect the acquisition of high-precision ground-based observation data for ground-based lunar radiation instruments. Among them, the tracking camera and signal detection lens being on the same optical axis are the key factors affecting the accuracy of the test, and the low-light radiation calibration coefficient is the key factor affecting the calibration accuracy. The tracking camera's full field of view of the moon's direct irradiance meter developed in this paper is 2.13°×1.60°, and the signal detection lens's full field of view is 1°. Because both fields of view are small, it is necessary to strictly parallel the optical axis of the tracking camera and the signal detection lens. If they are not adjusted to the same optical axis, it will cause the tracking camera to track the moon in real-time, the moon is at the center of the tracking camera's field of view but not in the center of the detection lens's field of view, resulting in no signal being detected or inaccurate data due to the signal being too weak. To ensure the accuracy of test data, this paper combines a laser level instrument with a flat reflection mirror to simulate the moon in the laboratory and adjust the tracking camera and detection lens on the same optical axis, avoiding inaccurate data caused by not being on the same optical axis. The experimental results show that the error of the same optical axis adjustment in the laboratory is within ±0.03°. In addition, the visible, infrared 1, and infrared 2 modules transmit radiation calibration, determining the coefficient relationship between the detector's output Digital Numbers(DN) value and the irradiance, which also directly affects the establishment of the lunar radiation model. To ensure the accuracy of the instrument's calibration coefficient, this paper uses a standard lamp traced back to the national metrology benchmark to transmit the calibration of the low-light lamp, using a combination of a laser rangefinder, a laser level instrument, and a flat reflection mirror to ensure the accuracy of the transmission calibration and solve the problems of non-linear response errors of the detector and long calibration distance errors caused by direct use of the standard lamp. In theory, this can improve the accuracy of low-light calibration. This paper aims to provide a universal and feasible solution to solve the problem of limited same optical axis adjustment due to environmental conditions affecting external field adjustment and to solve the problem of transmission radiation calibration accuracy and detector non-linear response using low-light sources for the moon's direct irradiance meter in the laboratory. This solution provides a way of thinking and a solution for laboratory optical axis adjustment and transmission calibration of integrated automatic observation remote sensing instruments with tracking and testing capabilities, and has important reference value.