Nowadays pre- and post-event disaster images are available in very high resolution by aerial imagery. Detecting buildings from aerial image in urban areas is a challenging task. Besides, finding equivalent buildings in pre- and post-disaster images is more challenging and hard to acquire. Accurate building detection in pre- and post-disaster images play significant role in damage assessment algorithms. This issue brought us to develop a common space for describing image pairs. To bate the problem, a method using PPI (Pixel Purity Index) and ISRA (Image Space Reconstruction Algorithm) is proposed. Some endmembers are extracted from pre-disaster image using PPI algorithm and both images are described according to these endmembers by ISRA. Our method could be a part of many conventional building detection algorithms and enable them being used in damage assessment algorithms. We embedded the proposed method into recently building detection algorithms to demonstrate the performance. It shows significant improvement in pre- and post-disaster images. It also shows that the purpose of finding equivalent buildings in pre- and post-disaster images is fully satisfied.
Baleghi, Y., & Kakooei, M. (2018). Building detection by common color space creation in pre- and post-disaster images. Electronics Industries, 9(3), 35-46.
MLA
yasser Baleghi; Mohammad Kakooei. "Building detection by common color space creation in pre- and post-disaster images". Electronics Industries, 9, 3, 2018, 35-46.
HARVARD
Baleghi, Y., Kakooei, M. (2018). 'Building detection by common color space creation in pre- and post-disaster images', Electronics Industries, 9(3), pp. 35-46.
VANCOUVER
Baleghi, Y., Kakooei, M. Building detection by common color space creation in pre- and post-disaster images. Electronics Industries, 2018; 9(3): 35-46.
)