A distributed edge extraction method for remote sensing images considering the completeness of results

A distributed edge extraction method for remote sensing images considering the completeness of results

Cheng Miaofang Chen Ken

Aiming at the problem of image feature loss caused by unreasonable data set partition in the edge extraction method of the existing distributed image, an elastic image pyramid (resilient image Pyramid, RIP) model is proposed, which guarantees the integrity of original image features by using multi-layer redundancy data segmentation, The pyramid is stored through key metadata to reduce the amount of data migration during the calculation. On this basis, a MapReduce calculation method for large remote sensing images is designed and implemented, and distributed edge extraction is carried out through Hadoop. The experimental results show that the extraction rate of image feature extracted by this method is improved, which is an effective method to guarantee the integrality of extraction results.

A distributed edge extraction method for remote sensing images considering the completeness of results