摘要
传统方式的舰船图像检索能够完成舰船图像的检索任务,但针对大规模复杂度较高的舰船图像,存在检索速度慢、检索不准确的问题,为此提出反馈机制的大规模舰船图像检索。利用图像反馈模型对舰船图像的颜色、纹理、形状、空间特征进行量化反馈,采用量化数值对比的方式进行图像检索;优化图像相似度度量方法,实现反馈机制的大规模舰船图像检索。实验数据表明,该方法比传统舰船图像检索速度提高66.44%,且具有良好的检索准确性。
Traditional way of ship image retrieval can complete ship image retrieval task,but in view of the complexity of mass higher ship image,has the problem of inaccurate retrieval slow and retrieval,image retrieval feedback mechanism is put forward for this large ships.Based on the color,texture,shape and spatial characteristic feedback value of ship image,quantization is carried out,and the image retrieval is carried out by quantitative comparison.Optimize image similarity measure,improve retrieval ability,realize the feedback mechanism of large-scale ship image retrieval.Experimental data show that the method is 66.44% faster than traditional ship image retrieval,and has good retrieval accuracy.
出处
《舰船科学技术》
北大核心
2018年第4X期157-159,共3页
Ship Science and Technology
基金
江西省教育厅科学技术研究项目(GJJ171048)
关键词
反馈机制
大规模检索
舰船图像
图像检索
feedback mechanism
large-scale retrieval
ship image
image retrieval
