Fire image processing and analysis technology plays a vital role in fire monitoring,fire accident rescue,and fire investigation.The precision of fire image analysis is closely related to the quality of the fire image to be processed,especially in the context of high noise pollution background,it is particularly important to improve the fire image quality.During image processing,the low-rank matrix recovery technology can explore its low-rank characteristics by itself,reducing the impact of noise intensity to some extent and showing excellent denoising characteristics.According to this,the paper regarding the video frame data of real-life fire video images in real cases,superimposes high-noise impact factors,and constructs a prior model of low-rank matrix by using the low-rank property of images.Through algorithm iteration,the low rank matrix approximation is obtained to establish a restoration model for high noise fire images.Comparing with traditional median filtering denoising method,the denoising effect of the low-rank matrix recovery method is better,which can better preserve image edge information and has a good effect on improving the accuracy of high noise fire image processing,thereby helping to improve the level of fire monitoring,command decision-making,and emergency rescue capabilities.
关键词
低秩矩阵恢复/图像去噪/火灾影像
Key words
low-rank matrix recovery/image denoising/fire scene images
维普
万方数据