Abstract
Depth sensing cameras (e.g., Kinect sensor, Tango phone) can acquire colorand depth images that are registered to a common viewpoint. This opens thepossibility of developing algorithms that exploit the advantages of bothsensing modalities. Traditionally, cues from color images have been used forobject localization (e.g., YOLO). However, the addition of a depth image can befurther used to segment images that might otherwise have identical colorinformation. Further, the depth image can be used for object size(height/width) estimation (in real-world measurements units, such as meters) asopposed to image based segmentation that would only support drawing boundingboxes around objects of interest. In this paper, we first collect color camerainformation along with depth information using a custom Android application onTango Phab2 phone. Second, we perform timing and spatial alignment between thetwo data sources. Finally, we evaluate several ways of measuring the height ofthe object of interest within the captured images under a variety of settings.