A new binary encoding algorithm for the integration of hyperspectral data and DSM

A new binary encoding algorithm for the integration of hyperspectral data and DSM is proposed in this paper. If the hyperspectral data consists of L spectral channels, the ordinary binary spectral encoding method represents the spectral amplitude and the spectral slope with a 2L-bit binary code vector. Usually the Hamming distance is chosen as the similarity measure to determine the spectral signature matches. Thinking that multi-source data may enhance the comprehension of images, we therefore attempt to integrate the spectral, shape and height information of remote sensing data of the same area by a modification of the binary encoding spectral matching method. Several processing steps are required before the integration. First, segments have to be established. The ground objects were derived from DSM and ground segments were established using HyMap image by an edge-based segmentation algorithm. Second, the mean spectrum was selected as the representative spectrum of the segment. After that, six shape descriptors, area, asymmetry, elliptic fit, rectangular fit, the ratio of length to width, and compactness, were calculated for each segments and transformed to binary codes. The relative heights derived from DSM were converted to binary codes too. The test data set covers an area in Oberpfaffenhofen, Germany. We tested the proposed method and the results show that the modified binary encoding classification is beneficial especially for discriminating similar spectral signatures.