A Comparison of 4 Shadow Compensation Techniques for Land Cover Classification of Shaded Areas from High Radiometric Resolution Aerial Images

作者：Wu Shou Tsung; Hsieh Yi Ta; Chen Chaur Tzuhn; Chen Jan Chang^*

来源：Canadian Journal of Remote Sensing, 2014, 40(4): 315-326.

DOI：10.1080/07038992.2014.979488

Very high resolution imagery offers great possibilities for land cover/use mapping. Unfortunately, very high resolution imagery leads to significant shadowy pixels. In recent years, aerial image devices have produced high radiometric resolution data (12-bit or higher), providing more radiometric detail of potential use in classification or interpretation of land cover of shadow areas. This study evaluated ADS-40 high radiometric resolution aerial images to determine the feasibility of classifying shadow areas. We analyzed the spectral properties of land cover in shadow areas and conducted shadow-image classification comparing 4 shadow compensation methods: Method 1, no treatment, used 13-bit spectral information in shaded areas for classification; Method 2 used linear correlation correction (LCC) before the classification; Method 3 used histogram matching (HM) before the classification; and Method 4 used multisource data fusion (MSDF) to aid in classification of shadows. Subsequently, we developed a practical proposal for the processing of shadow areas. The results of the 4 shadow processing methods of image classification indicated that Method 1 and Method 2 produced good results (Method 1, overall accuracy = 90.33%; Method 2, overall accuracy = 92.33%). Therefore, the high radiometric resolution images show potential for the classification of shadow areas.

• 出版日期2014