Due to the whitened or darkened CCD camera images of high-temperature objects with wide temperaturechange range, the object features in the images are weakened. In order to solve this problem, an automatic lightadjustment method based on average gray value for color CCD camera is proposed. In the investigation, first, theeffects of object heat radiation on the image gray during the imaging are discussed. Next, the image features of theobject with the temperature of 8501200℃are analyzed. Then, the relationship between the average gray of RGBprimary color images and the integral time is revealed. Finally, an automatic light adjustment method is constructed,which calculates the temperature zone and optimal integral time by using a prediction image frame, acquires thecolor image according to the optimal integral time, and extracts a proper primary gray image from the color imagebased on temperature zone for further usage. Experimental results show that the proposed method is robust to widetemperature change range and helps to acquire images with stable average gray value of high-temperature objects,which benefits the geometry measurement and surface defect detection of high-temperature objects.