首页|基于图像的雨强等级监测算法研究

基于图像的雨强等级监测算法研究

扫码查看
基于人工智能和深度学习技术,文章对不同雨强等级图片进行监测,提升了社会化摄像头对各雨强等级的监测能力.不同雨强等级的图像在HSV色彩空间、对比度、纹理特征、白色占比和灰度共生矩阵等特征上存在一定的差异,基于这五个特征,首先采用SVM+特征的方法进行雨强等级监测,但验证发现该方法监测效果不佳.其次,鉴于深度学习方法能够提取到图像中较多深层次特征,继而采用基于ResNet的深度学习方法进行雨强等级的监测.实验结果表明,深度学习方法在雨强等级监测中的准确率明显高于SVM+特征的方法.
Research on Rain Intensity Level Monitoring Algorithm Based on Image
Based on Artificial Intelligence and Deep Learning technology,this paper monitors images of different rain intensity levels to improve the ability of social cameras to monitor each rain intensity level.There are some differences in HSV color space,contrast,texture feature,white proportion,Gray Level Co-occurrence Matrix and other features between images with different rain intensity levels.Based on these five features,firstly,the SVM+feature method is used to monitor the rain intensity level,but it is found that the monitoring effect of this method is not good.Secondly,in view of the Deep Learning method can extract more deep features in the images.Then,the Deep Learning method based on ResNet is used to monitor the rain intensity level.The experimental results show that the accuracy of the Deep Learning method in rain intensity level monitoring is significantly higher than that of the SVM+feature method.

Deep Learningfeature extractionrain intensity level monitoringSVM

张志明、熊辉

展开 >

四川职业技术学院,四川 遂宁 629099

深度学习 特征提取 雨强等级监测 SVM

2024

现代信息科技
广东省电子学会

现代信息科技

ISSN:2096-4706
年,卷(期):2024.8(21)