The invention provides a bionic robotic fish interaction control method and system based on gesture recognition. According to the scheme, the bionic robotic fish interaction control method comprises the steps that gestures are obtained, wherein gesture sequence images are obtained through a visual sensor and the like; gesture recognition: through skin color detection, segmenting a gesture image; gesture features are extracted, a gesture library is established, and an SVM model is trained for gesture recognition; and gesture control: encoding a plurality of gesture actions to serve as a controlinstruction of the bionic robotic fish, and driving the robotic fish to move according to preset actions after gesture recognition is completed.
本公开提供了一种基于手势识别的仿生机器鱼交互控制方法及系统,所述方案包括获取手势:由视觉传感器等获取手势序列图像;手势识别:经过肤色检测,分割出手势图像;提取手势特征,建立手势库,训练SVM模型进行手势识别;手势控制:对若干手势动作进行编码,作为仿生机器鱼的控制指令,当手势识别完成之后,驱动机器鱼按照预定动作运动。
Bionic robotic fish interaction control method and system based on gesture recognition
一种基于手势识别的仿生机器鱼交互控制方法及系统
11.12.2020
Patent
Elektronische Ressource
Chinesisch