Abstract: To make full use of?the depth information for RGB-D image recognition, this paper proposes a new object recognition method based on tensor decomposition. Firstly, it represents the RGB-D image as a fourth-order tensor. Then, it decomposes the fourth-order tensor into a core tensor and four factor matrices. Finally, after projecting the fourth-order tensor by factor matrices, the newly obtained tensor is sent to a convolution neural network for object recognition. Comparative experimental results of three group similar objects on RGB-D dataset show that the proposed method obtains higher recognition accuracy than method that no-tensor fusing. Moreover, the single-object recognition accuracy can be improved by up to 99%.

Key words: RGB-D image fusion, convolutional neural network, tensor decomposition, Tucker decomposition, object recognition

摘要： 为了充分利用RGB-D图像的深度图像信息，提出了基于张量分解的物体识别方法。首先将RGB-D图像构造成一个四阶张量，然后将该四阶张量分解为一个核心张量和四个因子矩阵，再利用相应的因子矩阵将原张量进行投影，获得融合后的RGB-D数据，最后输入到卷积神经网络中进行识别。RGB-D数据集中三组相似物体的识别结果表明，利用张量分解融合RGB-D图像的物体识别准确率高于未采用张量分解的物体识别准确率，并且单一错分实例的准确率最高可提升99%。

关键词: RGB-D图像融合, 卷积神经网络, 张量分解, Tucker分解, 物体识别