Based on the study of the principle and the method of traditional camera calibration, a special template-matching-based monocular camera calibration method was used in the CCD-based automatic bacteria identification system, trying to improve the shortcomings of the traditional methods. This method directly used the 96-well elisa plate template in the system as a calibration auxiliary device. It only required the camera in the automatic microbiological identification system to capture a certain frame of the template image, and then it could calibrate the effective parameters such as the focal length and the height of the CCD camera in the system according to the constraints of the projective geometry. The whole progress required less calculation and was more practical. This method was proved to be more effective by comparing it with the spatial triangle calibration algorithm.