The unmanned combine harvester was developed using GPS, IMU, and machine vison.Especially, machine vison is the key to determine the working path by recognizing the column of rice.However, it has caused a fatal error for detecting the working path that paddy field has slope and uneven path unlike conventional vehicle road.Therefore, image calibration algorithm in accordance with the combine slope was needed.Image calibration system was composed of machine vision, stepping motor and inclinometer.Image calibration plate was constructed with the calibration line of the vertical and horizontal distance of 25 cm.Image calibration algorithm was developed using tilted plate image and value of inclinometer.The result shows that calibration line was tilted relative to center of camera at no slope condition.At slope condition of side direction,the calibration line was rotated clockwise or counter-clockwise relative to center of camera.At slope condition of front or rear direction, the calibration line was moved down or up relative to center of camera.The validation accuracy was about 95% at each slope condition.However, the validation accuracy was about 90% when the slope condition of front and rear direction and side direction have occurred at the same time.