基于空间金字塔池化特征的日常工具分类识别

Household tools classification recognition based on spatial pyramid pooling features

面向人机共融环境下机器智能对工具认知的需要,为提高家庭服务机器人的工具功能用途认知能力,设计一种基于深度几何特征空间金字塔池化的工具功用性建模与分类方法。离线训练阶段,考虑到各类工具在几何形态上的差异对工具自身更具表征性,首先,在各工具的深度图上提取多类几何特征,并融合形成工具特征图;然后,在工具特征图上提取多尺度特征块,并基于cciPCA的方法建立空间池化金字塔,从而构建最终的工具特征向量;最后,在高层语义空间上,利用SVM分类器训练工具分类识别模型。在线检测阶段,利用离线训练的工具分类模型对空间池化的样本进行分类测试。实验结果表明,所提方法能够实现家庭服务机器人对家庭日常工具的认知及分类识别,部分工具的识别精度可达97%及以上。

In view of the need of machine intelligence for tool recognition in man-machine communion environment, in order to improve the ability for home service robot to recognize affordances of tools, a modeling and classification method based on depth geometry feature spatial pyramid pooling is designed. In the off-line training phase, considering that the differences in the geometric shapes of various tools are more indicative for the tools themselves, firstly, the multi-types of geometric features are extracted from each tool depth map and fused to form the tool feature map. Then, the multi-scale feature patches are extracted and the spatial pooled pyramid is built based on cciPCA to construct the final tool feature vector. Finally, in the high-level semantic space, the tools classification model is trained based on the SVM classifier. In the online testing phase, the samples of spatial pooling are tested by a tool model trained offline. Experimental results show that the proposed method realizes recognition and classification of daily tools for home service robots, and the recognition accuracy of some tools reaches 97% and above.