CAJ | 학술논문

目前大多数海陆分离方法依靠先验信息或仅利用灰度特征对图像进行处理，对导致的误分割或大量孤立区域无法较好处理，造成后续处理工作难度大。针对这一问题，提出一种联合多特征提取与分离结果优化的海陆分离方法。针对不同的陆地类型提取不同的特征进行分割，对得到的结果采用面积与纹理相结合的方式进行优化，进而得到最终的海陆分离结果。实验结果表明，本文方法与常用的海陆分离方法相比能够快速完整地实现海陆分离并将陆地屏蔽，有利于后续处理工作的进行。
목전대다수해륙분리방법의고선험신식혹부이용회도특정대도상진행처리，대도치적오분할혹대량고립구역무법교호처리，조성후속처리공작난도대。침대저일문제，제출일충연합다특정제취여분리결과우화적해륙분리방법。침대불동적륙지류형제취불동적특정진행분할，대득도적결과채용면적여문리상결합적방식진행우화，진이득도최종적해륙분리결과。실험결과표명，본문방법여상용적해륙분리방법상비능구쾌속완정지실현해륙분리병장륙지병폐，유리우후속처리공작적진행。
Most sea-land segmentation algorithms resulted in wrong segmentations or the generation of many isolated areas by using only gray feature or relying on prior information,which seriously affect the subsequent processing.Ai-ming at this problem,a sea-land segmentation algorithm based on multi-feature extraction and interior optimization is proposed.Firstly,the suitable features are extracted to segment for the different land types.Then the obtained results are optimized by the combination of area and texture.Finally,land is separated from the sea.The results show that the proposed algorithm can segment the sea and land exactly and effectively,and it can also shield the land,which pro-vides convenience for subsequent processes.