广西大学学报（自然科学版）
廣西大學學報（自然科學版）
엄서대학학보（자연과학판）
JOURNAL OF GUANGXI UNIVERSITY （NATURAL SCIENCE EDITION）
2014年
6期
1285-1290
,共6页
摄像机标定%极限学习机%粒子群优化算法%双目视觉
攝像機標定%極限學習機%粒子群優化算法%雙目視覺
섭상궤표정%겁한학습궤%입자군우화산법%쌍목시각
camera calibration%extreme learning machine%particle swarm optimization%binocular vision
针对极限学习机( extreme learning machine,ELM)在隐层节点数较少时标定精度较低的问题,利用粒子群优化算法( particle swarm optimization,PSO)与极限学习机相结合的方法对双目视觉摄像机进行标定。在标定过程中,ELM直接描述图像信息与三维信息之间的非线性关系,然后利用PSO优化ELM的输入权值与隐层阈值。实验结果表明,与ELM相比较,基于粒子群极限学习机( PSO-ELM)的双目视觉摄像机标定方法能仅用较少隐层节点数获得较高精度。
針對極限學習機( extreme learning machine,ELM)在隱層節點數較少時標定精度較低的問題,利用粒子群優化算法( particle swarm optimization,PSO)與極限學習機相結閤的方法對雙目視覺攝像機進行標定。在標定過程中,ELM直接描述圖像信息與三維信息之間的非線性關繫,然後利用PSO優化ELM的輸入權值與隱層閾值。實驗結果錶明,與ELM相比較,基于粒子群極限學習機( PSO-ELM)的雙目視覺攝像機標定方法能僅用較少隱層節點數穫得較高精度。
침대겁한학습궤( extreme learning machine,ELM)재은층절점수교소시표정정도교저적문제,이용입자군우화산법( particle swarm optimization,PSO)여겁한학습궤상결합적방법대쌍목시각섭상궤진행표정。재표정과정중,ELM직접묘술도상신식여삼유신식지간적비선성관계,연후이용PSO우화ELM적수입권치여은층역치。실험결과표명,여ELM상비교,기우입자군겁한학습궤( PSO-ELM)적쌍목시각섭상궤표정방법능부용교소은층절점수획득교고정도。
If the number of hidden neuron is less, the calibration precision is low in extreme learn-ing machine( ELM) . Aiming at this problem, a method based on extreme learning machine opti-mized by particle swarm optimization ( PSO) is used in binocular camera calibration. The ELM neu-ral networks are used to describe the non-liner relation between the stereo image point and the 3 D geometry in the binocular vision while the PSO is used to optimize the input weights and the hidden biases in the ELM neural networks. Compared with ELM calibration method, experimental results show that the proposed binocular camera calibration method based on PSO-ELM could get a higher precision with less hidden neurons.
