Air pollution is one of the most important problems in the new era. Detecting the level of air pollution from an image taken by a camera can be informative for the people who are not aware of exact air pollution level be declared daily by some organizations like municipalities. In this paper, we propose a method to predict the level of the air pollution of a location by taking an image by a camera of a smart phone then processing it. We collected an image dataset from city of Tehran. Afterward, we proposed two methods for estimation of level of air pollution. In the first method, the images are preprocessed and then Gabor transform is used to extract features from the images. At the end, two shallow classification methods are employed to model and predict the level of air pollution. In the second proposed method, a Convolutional Neural Network(CNN) is designed to receive a sky image as an input and result a level of air pollution. Some experiments have been done to evaluate the proposed method. The results show that the proposed 9 method has an acceptable accuracy in detection of the air pollution level. Our deep classifier achieved accuracy about 59.38% which is 10 about 6% higher than traditional combination of feature extraction and classification methods. 相似文献
Geotechnical and Geological Engineering - Soil nailing is an in-situ soil reinforcement technique that is used to enhance the stability of land slopes, retaining walls and excavations. This... 相似文献
Theoretical and Applied Climatology - Snow is a key element for many socioeconomic activities in mountainous regions. Due to the sensitivity of the snow cover to variations of temperature and... 相似文献
Occurrence of drought, as an inevitable natural climate feature, cannot be ceased while happening. However, costs of the consequences could be alleviated using mature scientific integrated approaches. To reduce the amount of damage, it is required to provide “Contingency” and “Mitigation” action plans. For this reason, development of efficient operating instructions for various regions based on weather conditions and field studies is needed as well as having a sophisticated understanding of socioeconomic situations. This paper describes an approach to provide the first national agricultural drought risk management plan for a river basin in Iran country as a pilot. The study lasted for 3 years as a national technical research project for the “soil conservation and watershed management research institute.” To reach the objectives, besides holding workshops and specialized think-tank meetings, field researches were done. Based on the socioeconomic data sources in the basin and the results of meetings by participation of local managers and residents, the final plan was developed. Moreover, in order to carry out this research, different climatic, agricultural and local information were collected in the watershed. In the next steps, potential risks and vulnerabilities of various agricultural sectors due to the hazard were evaluated. In this study, a nine-step approach to develop an agricultural drought risk management plan proposing different scientific–managerial phases based on the latest experts’ opinions, released international scientific best practices, and existing conditions governing the region was followed. With respect to the average income of US$ one million from agriculture and animal husbandry in the river basin, total drought loss varies from US$ 86,000 to US$ 258,000 for a range of light to very intense drought conditions, respectively. The setup of these nine executive phases defined monitoring, forecasting, and warning steps in working teams and managed the subprograms in partnership with stakeholders and decision-makers to mitigate the rate of drought damage from 30 to 47% (depending on the severity of the drought condition).
This paper presents a u‐p (displacement‐pressure) semi‐Lagrangian reproducing kernel (RK) formulation to effectively analyze landslide processes. The semi‐Lagrangian RK approximation is constructed based on Lagrangian discretization points with fixed kernel supports in the current configuration. As a result, it tracks state variables at discretization points while allowing extreme deformation and material separation that is beyond the capability of Lagrangian formulations. The u‐p formulation following Biot theory is incorporated into the formulation to describe poromechanics of saturated geomaterials. In addition, a stabilized nodal integration method to ensure stability of the domain integration and kernel contact algorithms to model contact between bodies are introduced in the u‐p semi‐Lagrangian RK formulation. The proposed method is verified with several numerical examples and validated with an experimental result and the field data of an actual landslide. 相似文献
The precision study of dark matter using weak lensing by large-scale structure is strongly constrained by the accuracy with which one can measure galaxy shapes. Several methods have been devised but none has demonstrated the ability to reach the level of precision required by future weak lensing surveys. In this paper, we explore new avenues to the existing 'Shapelets' approach, combining a priori knowledge of the galaxy profile with the power of orthogonal basis function decomposition. This paper discusses the new issues raised by this matched filter approach and proposes promising alternatives to shape measurement techniques. In particular, it appears that the use of a matched filter (e.g. Sérsic profile) restricted to elliptical radial fitting functions resolves several well-known Shapelet issues. 相似文献