Luojia-1 Nightlight Image Registration Based on Sparse Lights

When mosaicking adjacent nightlight images of a large area that lacks human activities, traditional registration methods have difficulty realizing the tie point registrations due to lack structural information. In order address this issue, study devises an easy-to-implement engineering solution allows for sparse light areas with high efficiency while guaranteeing accuracy in non-sparse areas. The proposed method first extracts sparsely distributed positions through use roundness detection and centroid method. Then, geometric positioning forward backward algorithms random consistency sampling algorithm (RANSAC) are used achieve rough remaining points expanded affine model. Through experimentation it was found that, compared methods, is more reliable has wider distribution Finally, test 275 scenes China from Luojia-1, coverage ratio increased 59.3% 95.3% block adjustment 0.63 pixels, which verifies effectiveness provides basis registration, adjustment, images.