Automatic identification of illegal construction and demolition waste landfills: A computer vision approach.

Dozens of landslide accidents are reported at construction and demolition waste (CDW) landfills worldwide every year. Those accidents could be avoided via timely inspection in which the identification of illegal CDW landfills at a large scale plays a critical role. Traditional field surveys are time-consuming, labor-intensive, which is not effective in large-scale detection of landfills. To address this issue, a methodology is proposed in this study for the automatic identification of CDW landfills in large-scale areas by utilizing semantic segmentation of remote sensing imagery. Deep learning is employed to achieve automatic identification and a case study is conducted to showcase the models. The results shown that: (1) The model proposed in this study can effectively identify CDW landfills, with an accuracy of 96.30 % and an IoU of 74.60 %. (2) DeepLabV3+ demonstrated superior performance over Pspnet and HRNet, though HRNet approached DeepLabV3+ in performance with appropriate optimizations. (3) Case study results indicate the potential existence of 52 CDW landfills in Shenzhen, includng 4 official landfills and 48 suspected illegal CDW landfills, mainly in Longhua, Guangming, and Baoan districts. The method proposed in this study provides an effective approache to identify large-scale illegal CDW landfills and has great significance for supervising CDW landfills.

Assessing and predicting the illegal dumping risks in relation to road characteristics.

Using historical data to assess illegal dumping risks has significant potential to enhance the effectiveness of waste management in low-population density counties where the ability to patrol and regulate illegal dumping is limited. Using big data and geographical analysis to identify high-risk areas plays an important role in improving the effectiveness of supervision related to illegal dumping. However, current methods for classifying risk areas have limited accuracy. Taking an area in South Australia as an example, this study aims to improve the accuracy of classifying risk areas by using geo-information technology and machine learning methods. The results show that combining illegal dumping locations with road characteristics allows the high-risk areas to be refined to road sections. Compared with identifying the whole road or area as a high-risk spot, this result could be beneficial for monitoring illegal dumping in real life. Moreover, this model allows the analysis of factors that affect illegal dumping locations. Results show that the influencing factors for different risk levels of illegal dumping vary significantly. The model developed in this research can effectively distinguish risk levels according to these factors, and the model classification accuracy can reach 85%. In addition, there are priorities amongst these factors. This finding could help environmental authorities to allocate equipment and personnel with consideration of varying level of importance of those factors. This study has both technical contributions to identify high risk areas of illegal dumping, and theoretical implications for its management.

Status quo of illegal dumping research: Way forward.

Due to the rapid social and economic development, the past decades have witnessed the improvement of human being's quality of life and the speedy development of the construction industry. Meanwhile, the illegal dumping of solid waste has presented a significant issue. By using the method of systematic review, this study critically examined the literature related to illegal dumping that were published since 1990, and analyzed the current status and future trends of related research. Results show that the current studies on illegal dumping mainly focus on four perspectives: environmental science and toxicology, economics, management, and the use of emerging technologies. This critical review revealed that although the issue of illegal dumping has been widely recognized in recent years, some questions remain unanswered. Therefore, a future research agenda is proposed. These include: (1) Identifying the migration of pollutants in the food chain during the illegal dumping; (2) Implementing targeted treatment of illegal dumping pollutants; (3) Improving the stakeholder decision analysis model; (4) Expanding the scope of research on stakeholders of illegal dumping; (5) Formulating an unified evaluation standard for the related costs of illegal dumping; (6) Strengthening the evaluation of the interaction effects of influencing factors; (7) Comparing the effects of different types of factors; (8) the exploration of other influencing factors; (9) Analyzing illegal dumping by combining big data with the amount of solid waste; (10) Combining with monitoring to analyze the illegal dumping of household waste.