| Smoke concentration is the measure of the amount of smoke and the degree of smoke damage.It has important reference value in fire detection,fire evacuation and material safety assessment.The study of smoke concentration measurements is of great significance.In recent years,smoke concentration measurement method with video began to rise.In this paper,we first introduce the definition of image structure similarity index and its application in the qualitative detection of smoke.Then we apply the image structure similarity index to the smoke concentration quantitative measurement method.Calculate the image structure similarity index between the smoke images with different concentrations and smoke-free images in the video.Compare the results obtained with the smoke concentration values smoke meter measured in the same time.The results show that there is a certain quantitative relationship between the two values.By using the polyfit function in MATLAB,the first order polynomial fitting is carried out.According to the fitting polynomial,the conversion coefficient can be get by calculating smoke concentration according to the image structure similarity index under the background image.In addition,in order to make the method not limited by the background image,this paper extracts the texture features based on the gray covariance matrix.On this basis,we extract five characteristic parameters: the second moment,the contrast,the inverse moment,the entropy and the correlation.The weighted sum is taken as the value of the image complexity,and the conversion coefficient is replaced by the negative value of the image complexity in the formula.Therefore,if the complexity of the background image is calculated,we can calculate the smoke concentration according to the image structure similarity index in different background images.In this paper,different background images,different smoke concentration conditions were tested.The test results show that this method can quantitatively measure the exact value of smoke concentration in indoor conditions.At the same time,the method can also be applied to the detection of smoke at the beginning of the fire.The innovation of this paper is to propose the calculation formula of smoke concentration calculated by the image structure similarity index,and the conversion coefficient is replaced by the negative value of the image complexity in the formula.The quantitative measurement of smoke concentration based on similarity of image structure under any background images is realized.This method can not only qualitatively detect the existence of smoke,but also can accurately measure the specific value of smoke concentration.In addition,since the image structure similarity measures the loss of structure between the smoke images and the smoke-free image,the method will not be limited by the smoke color and the background brightness. |
