边缘检测和证据理论在机器人目标识别中的应用
|
摘要
以NAO机器人为平台,在机器人视觉算法的基础上,研究了NAO机器人的有色目标识别方法。采用基于HSI颜色空间的离散小波变换方式,利用可自动适应阈值Canny算子边缘检测方法和双阈值法提高了图像边缘的准确性和容错性。通过NAO机器人双摄像头系统进行图像的颜色标定,提出的基于模糊系统下证据理论的双摄像机融合策略能够有效地减少颜色之间的标定冲突,将目标图像的识别性能提高约29%。
On the NAO robot platform,this paper focuses on the NAO robot's colored target recognition method based on the research of robot vision algorithm. Using the discrete wavelet transform method based on HSI color space,the Canny operator edge detection method and double threshold method which can automatically adapt to the threshold are used to improve the accuracy and fault tolerance of image edges. Through the NAO robot dual camera system for color calibration of images,the proposed dual camera fusion strategy based on evidence theory in fuzzy system can effectively reduce the calibration conflict between colors and improve the recognition performance of target images by 29%.
On the NAO robot platform,this paper focuses on the NAO robot's colored target recognition method based on the research of robot vision algorithm. Using the discrete wavelet transform method based on HSI color space,the Canny operator edge detection method and double threshold method which can automatically adapt to the threshold are used to improve the accuracy and fault tolerance of image edges. Through the NAO robot dual camera system for color calibration of images,the proposed dual camera fusion strategy based on evidence theory in fuzzy system can effectively reduce the calibration conflict between colors and improve the recognition performance of target images by 29%.
引文
[1]刘胤,冯军焕,尹帮旭.图像识别技术在实验室监控中的应用[J].实验室研究与探索,2013,32(10):210-213.
[2] Jokinen K,Wilcock G. Multimodal open-domain conversations with the Nao robot[C]//Natural Interaction with Robots,Knowbots and Smartphones. NY:Springer,2014:213-224.
[3]李大华,赵相飞,许亮,等.基于纹理特征与HSI空间的苹果识别与标定[J].图学学报,2016,37(5):688-693.
[4]谢朝阳,李贵杰,彭忠明,等.基于证据理论和代理模型的QMU分析[J].电子科技大学学报,2018,47(1):66-72.
[5] English B A, Coates A, Howard A. Recognition of gestural behaviors expressed by humanoid robotic platforms for teaching affect recognition to children with autism-A healthy subjects pilot study[C]//International Conference on Social Robotics. Cham:Springer,2017:567-576.
[6]王伟,刘付显,徐浩.基于证据理论的群体研讨建模[J].系统工程与电子技术,2018(3):603-608.
[7] Mondou D,Prigent A,Revel A. A dynamic scenario by remote supervision:a serious game in the museum with a Nao robot[C]//International Conference on Advances in Computer Entertainment.Cham:Springer,2017:103-116.
[8]潘峥嵘,张玉阳,朱翔. NAO机器人自主目标跟踪算法研究[J].自动化与仪表,2017,32(8):12-16.
[9]王振,张星. NAO机器人平台的创新实验设计与实现[J].实验室研究与探索,2016,35(12):156-159.
[10]刘小峰,周旭,汪建明,等.大学生创新技能培养的NAO机器人交互平台[J].实验室研究与探索,2016,35(3):147-151.
[11]孙子文,李松,孙晓雯.基于DS证据理论的人体跌倒检测方法[J].计算机工程与科学,2018,40(5):829-835.
[12] Dong E Z,Wang D D,Tong J G,et al. A stable gait planning method of biped robot based on ankle motion smooth fitting[J].International Journal of Control,Automation and Systems,2018,16(1):284-294.
[13]刘标,许腾,李光. DS证据理论改进算法提高水下目标识别准确性[J].现代防御技术,2018(1):120-124.
[14] Yuan Q,Xi Z,Lu Q,et al. Method and experiment of the NAO humanoid robot walking on a slope based on CoM motion estimation and control[C]//International Conference on Intelligent Robotics and Applications. Cham:Springer,2017:154-165.
[15]周凤.一种HSI空间基于模糊聚类的火灾彩色图像分割算法[J].火灾科学,2017,26(1):49-53.
[2] Jokinen K,Wilcock G. Multimodal open-domain conversations with the Nao robot[C]//Natural Interaction with Robots,Knowbots and Smartphones. NY:Springer,2014:213-224.
[3]李大华,赵相飞,许亮,等.基于纹理特征与HSI空间的苹果识别与标定[J].图学学报,2016,37(5):688-693.
[4]谢朝阳,李贵杰,彭忠明,等.基于证据理论和代理模型的QMU分析[J].电子科技大学学报,2018,47(1):66-72.
[5] English B A, Coates A, Howard A. Recognition of gestural behaviors expressed by humanoid robotic platforms for teaching affect recognition to children with autism-A healthy subjects pilot study[C]//International Conference on Social Robotics. Cham:Springer,2017:567-576.
[6]王伟,刘付显,徐浩.基于证据理论的群体研讨建模[J].系统工程与电子技术,2018(3):603-608.
[7] Mondou D,Prigent A,Revel A. A dynamic scenario by remote supervision:a serious game in the museum with a Nao robot[C]//International Conference on Advances in Computer Entertainment.Cham:Springer,2017:103-116.
[8]潘峥嵘,张玉阳,朱翔. NAO机器人自主目标跟踪算法研究[J].自动化与仪表,2017,32(8):12-16.
[9]王振,张星. NAO机器人平台的创新实验设计与实现[J].实验室研究与探索,2016,35(12):156-159.
[10]刘小峰,周旭,汪建明,等.大学生创新技能培养的NAO机器人交互平台[J].实验室研究与探索,2016,35(3):147-151.
[11]孙子文,李松,孙晓雯.基于DS证据理论的人体跌倒检测方法[J].计算机工程与科学,2018,40(5):829-835.
[12] Dong E Z,Wang D D,Tong J G,et al. A stable gait planning method of biped robot based on ankle motion smooth fitting[J].International Journal of Control,Automation and Systems,2018,16(1):284-294.
[13]刘标,许腾,李光. DS证据理论改进算法提高水下目标识别准确性[J].现代防御技术,2018(1):120-124.
[14] Yuan Q,Xi Z,Lu Q,et al. Method and experiment of the NAO humanoid robot walking on a slope based on CoM motion estimation and control[C]//International Conference on Intelligent Robotics and Applications. Cham:Springer,2017:154-165.
[15]周凤.一种HSI空间基于模糊聚类的火灾彩色图像分割算法[J].火灾科学,2017,26(1):49-53.