Impact of Video Compression on the Performance of Object Detection Systems for Surveillance Applications
This addresses the trade-off between storage/bandwidth and detection accuracy for surveillance systems, but it is incremental as it applies an existing method to new data.
The study investigated how H.264 video compression affects YOLOv5 object detection performance in surveillance videos, finding that moderate compression (CRF 37 vs. 22) reduces file sizes without harming detection, but higher compression degrades it, especially in complex scenes, and retraining on compressed data improves F1 scores by up to 1% for highly compressed footage.
This study examines the relationship between H.264 video compression and the performance of an object detection network (YOLOv5). We curated a set of 50 surveillance videos and annotated targets of interest (people, bikes, and vehicles). Videos were encoded at 5 quality levels using Constant Rate Factor (CRF) values in the set {22,32,37,42,47}. YOLOv5 was applied to compressed videos and detection performance was analyzed at each CRF level. Test results indicate that the detection performance is generally robust to moderate levels of compression; using a CRF value of 37 instead of 22 leads to significantly reduced bitrates/file sizes without adversely affecting detection performance. However, detection performance degrades appreciably at higher compression levels, especially in complex scenes with poor lighting and fast-moving targets. Finally, retraining YOLOv5 on compressed imagery gives up to a 1% improvement in F1 score when applied to highly compressed footage.