Sensory Fusion and integration of LIDAR with stereo vision and Kinect for analysis of depth

Abstract

Sensory integration and fusion of a LIDAR system with two stereo vision cameras and Kinect for measuring depth is presented; filtering methods using maximum likelihood and Kalman for processing of data. it was developed a hardware fusion and later a software integration in order to carry out tests to 0,68, 1,5, 2, 2,5 and 4 m. Getting that the hardware fusion LIDAR-Kinect has the best result with an average error 1,08%, and in turn the Kalman filter provides better support for processing and accuracy of the data for both fusions with an average error of 1,35%.