| With the continuous development of sensor technology,the improvement of spatial resolution of remote sensing images brings richer spatial detail information,and the detection of buildings with high-resolution remote sensing images plays a key role in urban planning and other fields,while the detection of collapsed buildings based on post-earthquake remote sensing images helps to get rid of the dependence on pre-earthquake data,which is of great significance for timely emergency response.At present,building detection based on high-resolution remote sensing image is mainly divided into building detection in urban scenes and collapsed building detection in post-earthquake scenes.Although the former method based on morphological attribute profiles shows good performance in the absence of massive training samples by fine portrayal of buildings with multiple attributes and scales,it needs to break through the constraints of reasonable selection of attribute scales and conflicting evidence among attributes to establish a reliable unsupervised detection model.The latter gets rid of the reliance on preearthquake reference information,but ignores the challenges posed by the inaccessibility or absence of elevation information.To this end,this paper studies building detection and collapsed building identification methods from high-resolution remote sensing images,and the specific contents are as follows:(1)In the building detection method for urban scenes,a joint optimization and fusion building detection method for morphological attribute profiles(MAPs)is proposed.Firstly,the set of candidate building objects is first extracted through image segmentation and a set of discriminative rules.Secondly,the differential attribute profiles are filtered by a genetic algorithm,and an adaptive cross-probability genetic algorithm-differential attribute profiles(ACGA-DAPs)to extract potential building pixels.on this basis,an unsupervised decision fusion framework is established by constructing a novel statistics space building index(SSBI).Finally,the automated detection of buildings is realized.We show that the proposed method is significantly better than the state-of-the-art methods on high-resolution remote sensing images with different groups of different regions and different sensors,and overall accuracy(OA)of our proposed method is more than 91.9%.(2)In the collapsed building detection method in post-earthquake scenes,a detection method for collapsed buildings combining post-earthquake high-resolution optical and synthetic aperture radar(SAR)images is proposed by mining complementary information between traditional visual features and double bounce features from multi-source data.Firstly,a strategy of optical and SAR object set extraction based on an inscribed center(OSOIC)is first put forward to extract a unified optical-SAR object set.Based on this,a quantitative representation of collapse semantic knowledge in double bounce(QRCSD)is designed to bridge a semantic gap between double bounce and collapse features of buildings.At the same time,the underlying visual features of optical and SAR images are extracted based on the ACGA-MAPs proposed in the method of building detection in high-resolution remote sensing images,and the combined double bounce features and the underlying visual features of optical and SAR images form a feature space.Finally,the final detection results were obtained based on the improved active learning support vector machines(SVMs).The multi-group experimental results of post-earthquake multi-source images show that the OA and the detection accuracy for collapsed buildings(Pcb)of the proposed method can reach more than 82.39%and 75.47%.Therefore,the proposed method is significantly superior to many advanced methods for comparison. |
