Research on Semantic Object Measurement Algorithm Based on Object Detection

As a branch of machine vision, stereo vision has become a research hotspot in recent years. With the development of Augmented Reality (AR), image measurement technology has been proposed. We mainly study the algorithm of using RGB-D image to measure the size of objects in indoor environment, and comprehensively use computer vision technology and image target detection technology to realize the image measurement of objects. Aiming at the problem that there are many kinds of objects in the complex indoor environment and the measurement time in the actual environment, we propose a semantic target measurement algorithm based on target detection. Experimental results show that the target detection network we proposed can effectively identify the object category in indoor environment and measure the size of the recognized semantic target. At the same time. At the same time, the measurement time of the algorithm we proposed is very small, so the measurement efficiency can be improved.