In the process of becoming more complex and diverse society followed by economic and industrial development, remote sensing using satellite is on the rise in order to promptly grasp the change in land information. However, although there was the necessity to maintain traditional method of acquiring accurate information through digital map made through time and money consuming site investigation since low spatial resolution of previous satellite image is difficult to be used as reference data with low accuracy compared to measurement by site investigation, the range of its application is gradually becoming more diverse such as making of coastline map, digital map, and others as high resolution image with spatial resolution of under 1m was introduced with continuous advancement in remote sensing devices.
Digitizing process of current digital map manufacturing method which expresses the object through image recording and site measurement with naked eye is a method in which operator draws the object in person. Since it is manual and time and effort consuming method, it has disadvantage that the quality of result varies with the expertise of operator and type of software. Since errors may occur in subjective edition process of operator in such digitizing and vector edition process of digital map, it may bring about many problems in using it as reference data for accurate analysis of various studies. Therefore, it would have significance as reference data only when object is drawn through objective judgment instead of subjective method through operator in expressing the object to overcome such disadvantage.
Therefore, the algorithm for extraction of changed road network was proposed in this study for the renewal of road network on digital map through automatic extraction of road network using high resolution aerial orthophotography. In the first stage, road area was automatically extracted and error was removed through range expansion technique. In the second stage, changed area in road network was extracted through cva technique. In the third stage, the contour of road network extracted through edge filter was renewed to numerical map. At last, the accuracy of algorithm proposed in this study was examined through area comparison analysis of renewed road network.