Fast and accurate extraction of coastline is of great significance to the management of sea area. And object-oriented multi-scale segmentation method is used for automated extraction and classification coastlines from remote sensing imagery. Classification and extraction rule sets on coastal zone and coastline are set up according to their interpretation signs. Instantaneous waterline is extracted according to extraction rule sets; and a buffer zone to the inner land around this waterline is generated on the basis of extraction result; then coastal zone types are determined through classification. Artificial shoreline and bedrock shoreline are extracted firstly by their characteristics and the coastal zone classification results. Then coastal zone is re-segmented with artificial shoreline and bedrock coastline used as intervention mask, based on which sandy shoreline and developed mucky shoreline can be extracted. Tasseled cap transformation is applied to enhance the extraction result of vegetation on the non-developed muddy coastal beach, which can then be used to extract the non-developed muddy shoreline by rules sets. The experimental results show that the object-oriented classification and extraction method is effective for extraction of artificial shoreline, bedrock shoreline, sandy shoreline and muddy shoreline.
Image rectification is a common task in remote sensing application and usually time-consuming for large-size images.
Based on the characteristics of the Rational Functional Model (RFM)-based rectification process, this paper proposes a
novel CPU/GPU collaborative approach to high-speed rectification of remote sensing images. Three performance
optimization strategies are presented in detail, including maximizing device occupancy, improving memory access
efficiency and increasing instruction throughput. Experimental results using SPOT-5 and ZiYuan-3 (ZY3) remote
sensing images show that the proposed method can achieve the processing speed up to 8GB/min, which significantly
exceeds that of common commercial software. Real-time remote sensing image rectification can be expected with further
optimized algorithm and more efficient I/O operation.