The work involves the use of computer vision to enhance the accuracy of lighting elements detection in buildings. The aim is to improve the precision of the inventory of the location and state of lamps. The system has been improved with two modifications; a constraint on the orientation of the detected poses in the optimization methods, and an additional reprojection error filtering step to discard the erroneous poses introduced with the orientation restrictions. These improvements have been tested in five different case studies with over 30,000 images, showing significant enhancements in the number of detections and the percentage of correct model and state identifications.
Publication date: 19 Dec 2023
Project Page: Not provided
Paper: https://arxiv.org/pdf/2312.11380
