基于高斯混合模型的类人机器人果实辨识研究

Fruit Identification Research for Humanoid Robot Based on GMM Model Algorithm

针对类人机器人的特点,设计了一个具有果实辨识功能的类人机器人系统。类人机器人自带的摄像头作为视觉系统,通过视觉系统收集彩色果实图片,将RGB颜色转换成HSV颜色,结合高斯混合模型算法,通过训练得到每类果实所对应的分类器模型参数,构造分类器,实现类人机器人对果实准确高效的识别。

According to the characteristics of humanoid robot,a humanoid robot system of fruit identification was designed. Robot with type camera as vision,color fruit image was collected through the visual system. The RGB images will be converted into the HSV images,combined with the GMM model algorithm,the classifier model parameters were obtained by training,the classifier was constructed,so as to realize accurate and high efficient identification of fruits by humanoid robot.