Noncontact measurement of indoor objects with 3D laser camera-based

This paper proposes a novel noncontact measuring approach for indoor objects with 3D laser camera on the principle of time-of-flight. By Capturing the actual indoor sphere, the initial point cloud image is obtained and firstly filtered with statistical outlier removal and voxel grid removal methods to denoise outlier points and noises. The random sample consensus algorithm and Euclidean clustering are employed to segment others larger useless background. The sphere correction model is proposed to fit the actual size based on segmentation. Finally, compared to manual measurement results and under the distance of no more than two meters away the objects to be measured, the sizes of tens of indoor objects are measured and verified at the accuracy of 2%. This proposal could improve the remote measurement of moving target within two meters in case of inconvenient and dangerous proximity and will be employed to measure the body sizes of cattle at large-scale aquaculture at different breeding periods.