Abstract: In order to solve the problem of edge sawtooth in high-resolution image segmentation results, this study carries out high-resolution image contrast segmentation experiment with Landsat 8 multi-spectral data assisted or not in Huapiqiangzi forestry form in Yichun City, Heilongjiang Province. Firstly, the paper designs the segmentation experiment under the same scale parameter of multiresolution segmentation algorithm to determine the optimal composition of homogeneity criterion parameters required by the algorithm. Furthermore, based on the idea of local variance (LV) of object heterogeneity within a scene reflects the optimal segmentation scale, scales are found corresponding to obvious peaks of the homogenous local variance variation rate in the specific scale range (100~400, step size is 1) generated by ESP2, which is defined as the optimal segmentation scale range.Finally, the results of the two segmentation are evaluated by vector distance index, compactness index and shape index. The evaluation results show that compared with the independent segmentation of GF-2 image, the GF-2 image segmentation assisted by Landsat 8 multispectral data improves the quality of vector distance index, compactness index and shape index, and the average increase rate are 8.05%, 28.40%, 11.76% separately.

Key words: multiresolution segmentation, GF-2, Landsat 8, edge sawtooth, segmentation evaluation