Adaptive adjustment method and apparatus for fill light, and electronic device and storage medium
By dividing the supplementary lighting area in the monitoring equipment and using the supplementary lighting prediction model for feature fusion, the intensity of the supplementary lighting is adaptively adjusted, which solves the problems of light pollution and face overexposure caused by supplementary lighting and improves the image quality of the monitoring equipment.
WO2026144083A1PCT designated stage Publication Date: 2026-07-09ZHEJIANG UNIVIEW TECH CO LTD
Patent Information
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- ZHEJIANG UNIVIEW TECH CO LTD
- Filing Date
- 2025-07-03
- Publication Date
- 2026-07-09
Smart Images

Figure CN2025106756_09072026_PF_FP_ABST
Abstract
The present disclosure relates to the technical field of video monitoring. Provided are an adaptive adjustment method and apparatus for a fill light, and an electronic device and a storage medium. The method comprises: acquiring a face monitoring image collected by a monitoring device; inputting the face monitoring image into a fill-lighting prediction model, so as to obtain a target fill light combination result corresponding to the monitoring device, wherein the fill-lighting prediction model is obtained by means of performing training on the basis of calibration sample images and label data, each fill-lighting region is determined on the basis of a radiation angle corresponding to each group of fill lights in the monitoring device, and the target fill light combination result is obtained by means of the fill-lighting prediction model performing feature fusion on region images of the fill-lighting regions corresponding to the groups of fill lights in the face monitoring image; and on the basis of the target fill light combination result, performing adaptive adjustment on the intensity of each target fill light in the monitoring device.
Need to check novelty before this filing date? Find Prior Art