[19p-221B-1~10] 【CS.2】 Code-sharing Session of 3.3 & 4.4

▲ [19p-221B-9] Circular 3D Feature Descriptor for RGB-D Images

〇Jun Qiu¹, Lina Wu¹, Chang Liu¹ (1.Insti. of Applied Mathematics, Beijing Information Science and Technology Univ.)

Keywords:RGB-D image, 3D feature descriptor, rotational invariance

We proposed a circular descriptor of the 3D feature points of the RGB-D image by utilizing the rotation invariance.The circular feature descriptor with rotational invariance was proposed to describe the 3D local feature points of RGB-D images. Due to the computational complexity and the feature vector’s high dimension of conventional RGB-D descriptors, we investigated an efficient description method with circular computing property and rotational invariance. The proposed descriptor can be utilized to avoid the orientation assignment for the 3D local feature points and reduce the dimension of the feature vector.