數(shù)據(jù)配準(zhǔn)是多傳感器數(shù)據(jù)融合的先期處理過程。在研究了傳統(tǒng)的最小二乘算法的基礎(chǔ)上, 提出了基于三維坐標(biāo)系中的極大似然配準(zhǔn)算法, 解決了最小二乘配準(zhǔn)算法對于傳感器相距較遠(yuǎn)時(shí)無法解決的問題。最后通過計(jì)算機(jī)仿真驗(yàn)證了算法的有效性。
關(guān)鍵詞: 數(shù)據(jù)融合; 數(shù)據(jù)配準(zhǔn); 極大似然配準(zhǔn)法; 仿真分析
Abstract:Data regist rat ion is a p rerequ isite p rocess in m u li2sen so r data fu sion. O n the basis of wo rk s on t radit ional least squ ires techn iques, a m ax im um likelihood Regist rat ion A lgo rithm in 32D space is p resen ted in th is paper, that so lves the p rob lem that the least square regist rat ion algo rithm fails to handle w hen the sen so rs are far f rom each o ther. F inally, the eff iciency of th is algo rithm is demon st rated via compu ter sim u2
lat ion.
Key words: data fu sion, data regist rat ion. m ax im um likelihood regist rat ion algo rithm , sim u lat ion analy2 sis.