Visual Detection Of Construction Equipment Emission

3081 Words May 17th, 2015 13 Pages
Visual Detection Of Construction Equipment Emission
Yi Liu

Construction industry has great impact in the environment, especially in the pollutant gas emission. Measuring the emission is a vital part for us to take actions in order to enhance the compliance with environmental regulations and achieve the environmental goals. Most of the previous researches only focus on the operation phase of emission instead of the construction phase. But the fact is that more than 50% of the emissions come from the construction site equipment. In this paper, a new method has been presented to help solve the problem by visual monitoring the equipment activity and calculating the total emission based on the emission factor. In order to realize this, algorithm about Histograms of Oriented Gradients will be adapted and the Supported Vector Machine (SVM) classifier will be used to localize the equipment facture. After the geo-location of the features that has been generated in each frame, a decision tree can be adopted in order to classify the activity. The experiment results and potential benefit for the methods will be provided in detail. Then, the portable emission measurement system (PEMS) will be employed to estimate the total emissions.

1. Introduction
Monitoring the emission of the building is a vital part for the company and government to take action of decreasing the negative impact to the environment. But most of the previous studies only focus on the operational phase…
Open Document