Aiming at the difficulties in change detection caused by the complexity of high-resolution remote sensing images that exist in varied ecological environments and artificial objects, in order to overcome the limitations in traditional pixel-oriented change detection methods and improve the detection precision, an innovative object-oriented change detection approach based on multiscale fusion is proposed. This approach introduced the classical color texture segmentation algorithm J-segmentation (JSEG) to change detection and achieved the multiscale feature extraction and comparison of objects based on the sequence of J-images produced in JSEG. By comprehensively using the geometry, spectrum, and texture features of objects, and proposing two different multiscale fusing strategies, respectively, based on Dempster/Shafer evidence theory and weighted data fusion, the algorithm further improves the divisibility between changed and unchanged areas, thereby establishing an integrated framework of object-oriented change detection based on multiscale fusion. Experiments were performed on high-resolution airborne and SPOT 5 remote sensing images. Compared with different object-oriented and pixel-oriented detection methods, results of the experiments verified the validity and reliability of the proposed approach.