Method and device for detecting flicker of light source in video sequence and image processing apparatus

Abstract

The invention is applicable to the technical field of image and video processing, and provides a method and device for detecting the flicker of a light source in a video sequence and an image processing device. The method comprises the following steps of: acquiring a previous frame image and a current frame image; Extracting brightness information of a previous frame image and brightness information of a current frame image, and calculating a brightness line average value of the previous frame image and a brightness line average value of the current frame image; Calculating a difference function of a luminance line average value of a previous frame image and a luminance line average value of a current frame image, and integrating the difference function point by point to obtain a flicker characteristic component; Analyzing the scintillation characteristic component to obtain an analysis result; According to the analysis results and the preset flicker frequency threshold value, whetherthe current frame image has flicker phenomenon or not is judged. The invention can detect whether the flicker phenomenon caused by the artificial light source exists in the video sequence, has low time complexity and high detection accuracy, does not require the detected frame image to be a still frame image, and can be applied to the flicker detection of the light source in the moving video sequence.

Description

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine

Login to View More

AI Technical Summary

Benefits of technology

The technology described in this patented allows for quick identification of any lights or other sources causing visual disturbances during videos without requiring manual inspections on each camera's screen. It uses an algorithm called Flickering Difference Function (FLD) to calculate differences from both frames before and after they were captured with cameras. This results in points at specific locations where no visible changes have been observed due to natural illumination such as sunlight. By analyzing these points, we aimed towards identifying signs like flipping blinds over windows or shadows falling onto walls while driving cars.

Problems solved by technology

This patented technical problem addressed in this patents relates to improving the quality of captured cameras without causing blurriness during high ambient illumination conditions while maintaining consistence between darkest areas and brighter ones. Current solutions involve setting specific values beforehand every few seconds, which may result in unnatural visual effects like flashes overlapping across both fields.

Images