Land Use/Cover Classification by Using Digital Camera Imagery

This paper present an economical analysis of using a digital camera imagery data to classify land use/cover in the Prai Industrial area, Penang, located in Peninsular Malaysia. The data were captured using a digital camera, Kodak DC 290 from a small light aircraft at 8000 feet altitude. This overcomes the problem of difficulty in obtaining cloud-free satellite images especially in the equatorial region where experience showed that just two or three cloud-free scenes per year could be obtained. The use of digital camera as a sensor to capture digital images is cheaper and economical compared to the use of other airborne sensor. The images consisted of the three visible bands-red, green and blue. Three supervised classifications techniques (maximum likelihood, minimum distance-to-mean and parallelepiped) were performed to the digital image. The training sites were established using polygons within each scene and four land cover classes were assigned to each classifier. The relative performance of the techniques was evaluated. The accuracy of the classified images was validated using a reference data set. The results produced high degree of accuracy. Finally, geometric correction was performed to the digital image using the nearest neighborhood method with second order polynomial to produce geocoded map. The final results showed that the digital camera can be used as a tool for providing useful data for land cover classification. The classified images provided useful information for planning and development of a small area of coverage.