结合感受野模块与并联RPN网络的火焰检测  

Flame detection combined with receptive field and parallel RPN

作　　者：鲍文霞[1] 孙强 梁栋[1] 胡根生[1] 杨先军[2] Bao Wenxia;Sun Qiang;Liang Dong;Hu Gensheng;Yang Xianjun(School of Electronics and Information Engineering,Anhui University,Hefei 230601,China;Hefei Institutes of Physical Science,Chinese Academy of Sciences,Hefei 230031,China)

机构地区：[1]安徽大学电子信息工程学院,合肥230601 [2]中国科学院合肥物质科学研究院,合肥230031

出　　处：《中国图象图形学报》2023年第2期418-429,共12页Journal of Image and Graphics

基　　金：国家自然科学基金项目(61672032)。

摘　　要：目的准确快速的火焰检测技术在早期火灾预警中具有重要的实际应用价值。为了降低伪火类物体引起的误警率以及早期小火焰的漏检率,本文设计了一种结合感受野(receptive field,RF)模块与并联区域建议网络(parallel region proposal network,PRPN)的卷积神经网络(receptive field and parallel region proposal convolutional neural network,R-PRPNet)用于火焰检测。方法R-PRPNet主要由特征提取模块、并联区域建议网络和分类器3部分组成。特征提取模块在MobileNet卷积层的基础上,通过嵌入感受野RF模块扩大感受野捕获更丰富的上下文信息,从而提取更具鉴别性的火焰特征,降低伪火类物体引起的误警率;并联区域建议网络与特征提取模块后端的多尺度采样层连接,使用3×3和5×5的全卷积进一步拓宽多尺度锚点的感受野宽度,提升PRPN对不同尺度火焰的检测能力,解决火灾发生初期的小火焰漏检问题;分类器由softmax和smooth L1分别实现分类与回归。在R-PRPNet训练过程中,将伪火类物体作为负样本进行负样本微调,以更好区分伪火类物体。结果在包括室内、建筑物、森林和夜晚等场景火焰数据以及包括灯光、晚霞、火烧云和阳光等伪火类数据的自建数据集上对所提方法进行测试,在火焰检测任务中,准确度为98.07%,误警率为4.2%,漏检率为1.4%。消融实验结果表明,R-PRPNet较基线网络在漏检率和误警率上分别降低了4.9%和21.72%。与传统火焰检测方法相比,R-PRPNet在各项指标上均优于边缘梯度信息和聚类等方法。性能较几种目标检测算法有所提升,其中相较于YOLOX-L,误警率和漏检率分别降低了22.2%和5.2%。此外,本文在不同场景火焰下进行测试,都有较稳定的表现。结论本文方法有效降低了火焰检测中的误警率和漏检率,并可以满足火焰检测的实时性和准确性需求。Objective Early flame detection is essential for quick response event through minimizing casualties and damage.Smoky and flaming alarms have been using in indoor-scenario in common.However,the challenging issue of most traditional physical sensors is limited to the fire source-near problem and cannot be meet the requirement of outdoor-scene flame detection.Real-time image detection has been developing in terms of image processing and machine learning technique.However,the flame shape,size,and color can be varied intensively,and there are a plenty of pseudo-fire objects(very similar to the features of the flame color)in the natural environment.To distinguish real flames from pseudo flames precisely,the detection model has been developing dramatically.Image processing and machine learning methods can be divided into three categories:1)traditional image processing,2)machine learning,and 3)deep learning.Traditional image processing and machine learning is often concerned of design-manual of flame features,which is not quantitative and poor matching to complex background images.Thanks to the self-learning and deep learning techniques,current flame-based detection and interpretation has been facilitating.First,convolution-depth can be used to interpret small-scale areas less than 32 X 32 missing information on the feature map.Second,deep learning models can be applied to detect color features similar to the object-targeted and the misjudgment-caused,while it is restricted of small target and color-similar feature in flame-detection interpretation.In order to alleviate the pseudo-fire-objects-derived false alarm rate and the missed detection rate of early small flames,we develop a receptive field module based(RF-module-based)convolutional neural network(CNN)and the parallel region proposal network(PRPN)is designed for flame detection,called R-PRPNet.Method The R-PRPNet is mainly composed of three parts:1)feature extraction module,2)region-parallel network,and 3)classifier.The feature extraction module is focused on conv

关 键 词：火焰检测 深度学习 感受野(RF) 并联区域建议网络(PRPN) 负样本微调 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]