In this study, an image processing based volume measurement system is presented. The proposed scheme provides high accuracy, low cost and ease of calibration. There are no additional requirements such as alignment of objects during measurement, positioning the objects at a certain position in the measuring field, and these advantages reduce the labor effort in real-time. In the first stage, a problem-specific calibration process was performed and then Hough transform was applied in the image processing phase. Although the proposed method has been developed primarily to measure the volume of box-type objects, it also provides an approximate volume measurement for objects of random geometry. The system tested under various external disturbances has produced satisfactory results.