THE EFFECT OF LOSSY IMAGE COMPRESSION ON OBJECT BASED IMAGE CLASSIFICATION – WORLDVIEW-2 CASE STUDY

Abstract

Abstract. Lossy compression is becoming increasingly used in remote sensing although its effect on the processing results has yet not been fully investigated. This paper presents the effects of JPEG 2000 lossy compression on the classification of very high-resolution WorlView-2 imagery. The k-nearest neighbor and support vector machine methods of the object based classification were used and compared. The results explore the impact of compression on the images, segmentation and resulting classification. The study proves that in general lossy compression does not adversely affect the classification of images; what is more, in some cases classification of compressed images gives better results than classification of the original image. Classification accuracy of support vector machines method indicates that compression ratios of up to 30:1 can be used without any loss of accuracy. The best result of the k-nearest neighbor method was obtained with the highest compression ratio (100:1), but the outcome cannot be trusted without reserve. In the study we found that the support vector machine method gives better classification results than the k-nearest neighbor and is also recommended for further research. In addition to the classification method, image segmentation, a basic step of object classification, plays an important role in the accuracy of the results.