Add time:08/01/2019         Source:sciencedirect.com

In this paper, we propose a robot photography system that can autonomously search for the optimal target viewpoint. Two technical issues of scene composition evaluation and viewpoint selection are solved by this system. A composition evaluation method for photos is developed using well-known composition rules based on Kullback–Leibler (KL) divergence, considering the directional information of each target. To reduce the calculation cost of the composition evaluation in the case where the number of targets is large, automatic target grouping is conducted via variational Bayes. The optimal viewpoint with respect to the composition is selected from a number of candidate viewpoints around the targets based on KL divergence. Finally, the fact that better composed photos can be autonomously photographed by the proposed system is validated via experiments and human evaluations.

We also recommend Trading Suppliers and Manufacturers of Fast red KL base (cas 12216-86-3). Pls Click Website Link as below: cas 12216-86-3 suppliers