In the process of handeye calibration, the obtained data of the robot arm and the calibration board image have a great influence on the result of the handeye calibration. If the acquired data are not accurate, the calibration result may have a large error. Aimed at the problem that it is necessary to automatically control the robot arm to collect effective images in the process of online handeye calibration, an azimuth analysis method of calibration board based on block image histogram was presented. The method was used to block the image, which provided basis for obtaining the image azimuth information. The histogram of the block image was extracted, and the peak, peak position and halfwidth information were obtained according to the secondorder Gaussian fitting method. The peak position was compared with the threshold, and the peak area was solved to obtain the azimuth information of the calibration board, 〖JP2〗which provided guidance for the next movement of the robot arm. The experiment results show that by using the method, the azimuth of the calibration board can be accurately determined, and the efficiency and accuracy of calibration board image collection can be improved