Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
153 tokens/sec
GPT-4o
7 tokens/sec
Gemini 2.5 Pro Pro
45 tokens/sec
o3 Pro
4 tokens/sec
GPT-4.1 Pro
38 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Uncovering local aggregated air quality index with smartphone captured images leveraging efficient deep convolutional neural network (2308.03200v3)

Published 6 Aug 2023 in cs.CV and cs.AI

Abstract: The prevalence and mobility of smartphones make these a widely used tool for environmental health research. However, their potential for determining aggregated air quality index (AQI) based on PM2.5 concentration in specific locations remains largely unexplored in the existing literature. In this paper, we thoroughly examine the challenges associated with predicting location-specific PM2.5 concentration using images taken with smartphone cameras. The focus of our study is on Dhaka, the capital of Bangladesh, due to its significant air pollution levels and the large population exposed to it. Our research involves the development of a Deep Convolutional Neural Network (DCNN), which we train using over a thousand outdoor images taken and annotated. These photos are captured at various locations in Dhaka, and their labels are based on PM2.5 concentration data obtained from the local US consulate, calculated using the NowCast algorithm. Through supervised learning, our model establishes a correlation index during training, enhancing its ability to function as a Picture-based Predictor of PM2.5 Concentration (PPPC). This enables the algorithm to calculate an equivalent daily averaged AQI index from a smartphone image. Unlike, popular overly parameterized models, our model shows resource efficiency since it uses fewer parameters. Furthermore, test results indicate that our model outperforms popular models like ViT and INN, as well as popular CNN-based models such as VGG19, ResNet50, and MobileNetV2, in predicting location-specific PM2.5 concentration. Our dataset is the first publicly available collection that includes atmospheric images and corresponding PM2.5 measurements from Dhaka. Our codes and dataset are available at https://github.com/lepotatoguy/aqi.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (59)
  1. Brook, R. D. et al. Air pollution and cardiovascular disease: a statement for healthcare professionals from the expert panel on population and prevention science of the american heart association. \JournalTitleCirculation 109, 2655–2671 (2004).
  2. Human health effects of air pollution. \JournalTitleEnvironmental pollution 151, 362–367 (2008).
  3. Landrigan, P. J. Air pollution and health. \JournalTitleThe Lancet Public Health 2, e4–e5 (2017).
  4. Star, T. D. Dhaka again ranks world’s most polluted city. https://www.thedailystar.net/environment/pollution/air-pollution/news/dhaka-again-ranks-worlds-most-polluted-city-2976026 (2022). Accessed: 2022-3-29.
  5. Star, T. D. 2021 world air quality report: Bangladesh remains most polluted. https://www.thedailystar.net/environment/pollution/air-pollution/news/2021-world-air-quality-report-bangladesh-most-polluted-2987961 (2022). Accessed: 2022-3-29.
  6. Particle pollution estimation based on image analysis. \JournalTitlePLoS One 11, e0145955 (2016). DOI https://doi.org/10.1371/journal.pone.0145955.
  7. Highly efficient picture-based prediction of PM2.5 concentration. \JournalTitleIEEE Trans. Ind. Electron. 66, 3176–3184 (2019).
  8. LIBSVM: a library for support vector machines. \JournalTitleACM Trans. Intel. Syst. Technol 2 (2011).
  9. Auvee, M. S. B. Air pollution monitoring system based on air pollutant index(api) and geographic information system(gis). http://dspace.bracu.ac.bd/xmlui/bitstream/handle/10361/12290/17341043_CSE.pdf (2019). Accessed: 2022-3-29.
  10. The World Air Quality Index project. A beginner’s guide to air quality Instant-Cast and Now-Cast. https://aqicn.org/faq/2015-03-15/air-quality-nowcast-a-beginners-guide/ (2015). Accessed: 2023-9-19.
  11. A deep learning and image-based model for air quality estimation. \JournalTitleScience of The Total Environment 724, 138178 (2020). URL https://www.sciencedirect.com/science/article/pii/S0048969720316910. DOI https://doi.org/10.1016/j.scitotenv.2020.138178.
  12. Zhang, Q. et al. Real-time and image-based aqi estimation based on deep learning. \JournalTitleAdvanced Theory and Simulations 2100628 (2022). DOI https://doi.org/10.1002/adts.202100628.
  13. Image-based air quality analysis using deep convolutional neural network. In 2017 IEEE International Conference on Image Processing (ICIP), 3949–3952 (IEEE, 2017).
  14. Using user generated online photos to estimate and monitor air pollution in major cities. In Proceedings of the 7th International Conference on Internet Multimedia Computing and Service - ICIMCS '15 (ACM Press, 2015). URL https://doi.org/10.1145%2F2808492.2808564. DOI 10.1145/2808492.2808564.
  15. Single image haze removal using dark channel prior. In 2009 IEEE Conference on Computer Vision and Pattern Recognition, 1956–1963 (IEEE, 2009).
  16. Learning depth from single monocular images using deep convolutional neural fields. \JournalTitleIEEE Trans. Pattern Anal. Mach. Intell. 38, 2024–2039 (2016).
  17. Rijal, N. et al. Ensemble of deep neural networks for estimating particulate matter from images. In 2018 IEEE 3rd International Conference on Image, Vision and Computing (ICIVC), 733–738 (IEEE, 2018).
  18. Duong, D. Q. et al. Multi-source machine learning for AQI estimation. In 2020 IEEE International Conference on Big Data (Big Data) (IEEE, 2020).
  19. IMAGE-2-AQI: Aware of the surrounding air qualification by a few images. In Lecture Notes in Computer Science, 335–346 (Springer International Publishing, Cham, 2021).
  20. IoT-based AQI estimation using image processing and learning methods. In 2022 IEEE 8th World Forum on Internet of Things (WF-IoT) (IEEE, 2022).
  21. No-Reference image quality assessment in the spatial domain”. \JournalTitleIEEE Transactions on Image Processing (2012).
  22. Rath, S. YOLOv5 - fine tuning & custom object detection training. https://learnopencv.com/custom-object-detection-training-using-yolov5/ (2022). Accessed: 2023-9-17.
  23. Ahmed, M. et al. AQE-Net: A deep learning model for estimating air quality of karachi city from mobile images. \JournalTitleRemote Sens. (Basel) 14, 5732 (2022).
  24. Utomo, S. et al. AIX implementation in image-based PM2.5 estimation: Toward an AI model for better understanding. In 2023 15th International Conference on Knowledge and Smart Technology (KST), 1–6 (IEEE, 2023).
  25. Air quality prediction using CNN+LSTM-based hybrid deep learning architecture. \JournalTitleEnviron. Sci. Pollut. Res. Int. 29, 11920–11938 (2022).
  26. Hybrid interpretable predictive machine learning model for air pollution prediction. \JournalTitleNeurocomputing 468, 123–136 (2022).
  27. Wang, J. et al. An air quality index prediction model based on CNN-ILSTM. \JournalTitleSci. Rep. 12, 1–16 (2022).
  28. A deep learning approach for prediction of air quality index in a metropolitan city. \JournalTitleSustain. Cities Soc. 67, 102720 (2021).
  29. McCartney, E. J. Optics of the atmosphere: scattering by molecules and particles. \JournalTitleNew York (1976).
  30. Vision and the atmosphere. \JournalTitleInternational journal of computer vision 48, 233–254 (2002).
  31. Fattal, R. Dehazing using color-lines. \JournalTitleACM transactions on graphics (TOG) 34, 1–14 (2014).
  32. Improved single image dehazing using geometry. In 2009 Digital Image Computing: Techniques and Applications, 103–110 (2009). DOI 10.1109/DICTA.2009.25.
  33. Fattal, R. Single image dehazing. \JournalTitleACM transactions on graphics (TOG) 27, 1–9 (2008).
  34. Real-time air quality estimation based on color image processing. In 2014 IEEE Visual Communications and Image Processing Conference, 326–329 (2014). DOI 10.1109/VCIP.2014.7051572.
  35. Huang, W. et al. Visibility, air quality and daily mortality in shanghai, china. \JournalTitleScience of the Total Environment 407, 3295–3300 (2009).
  36. Human perception of visual air quality. \JournalTitleJournal of the Air Pollution Control Association 30, 122–131 (1980).
  37. Bold fmri and psychophysical measurements of contrast response to broadband images. \JournalTitleVision research 44, 669–683 (2004).
  38. Air cognizer: Predicting air quality with TensorFlow lite. https://blog.tensorflow.org/2019/02/air-cognizer-predicting-air-quality.html. Accessed: 2020-1-22.
  39. About air quality index (AQI). http://case.doe.gov.bd/index.php?option=com_content&view=article&id=9&Itemid=31 (2011). Accessed: 2022-4-14.
  40. Vaswani, A. et al. Attention is all you need. \JournalTitleAdvances in neural information processing systems (NeurIPS) 30 (2017).
  41. Dosovitskiy, A. et al. An image is worth 16x16 words: Transformers for image recognition at scale (2020). URL https://arxiv.org/abs/2010.11929. DOI 10.48550/ARXIV.2010.11929.
  42. Li, D. et al. Involution: Inverting the inherence of convolution for visual recognition. In 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) (IEEE, 2021).
  43. A feature-enriched completely blind image quality evaluator. \JournalTitleIEEE Trans. Image Process. 24, 2579–2591 (2015).
  44. Gu, K. et al. No-reference quality assessment of screen content pictures. \JournalTitleIEEE Trans. Image Process. 26, 4005–4018 (2017).
  45. A highly efficient method for blind image quality assessment. In 2015 IEEE International Conference on Image Processing (ICIP) (IEEE, 2015).
  46. Gu, K. et al. No-reference quality metric of contrast-distorted images based on information maximization. \JournalTitleIEEE Trans. Cybern. 47, 4559–4565 (2017).
  47. Learning a no-reference quality assessment model of enhanced images with big data. \JournalTitleIEEE Trans. Neural Netw. Learn. Syst. 29, 1301–1313 (2018).
  48. S3: a spectral and spatial measure of local perceived sharpness in natural images. \JournalTitleIEEE Trans. Image Process. 21, 934–945 (2012).
  49. A fast wavelet-based algorithm for global and local image sharpness estimation. \JournalTitleIEEE Signal Process. Lett. 19, 423–426 (2012).
  50. No-reference image sharpness assessment in autoregressive parameter space. \JournalTitleIEEE Trans. Image Process. 24, 3218–3231 (2015).
  51. Li, L. et al. No-reference image blur assessment based on discrete orthogonal moments. \JournalTitleIEEE Trans. Cybern. 46, 39–50 (2016).
  52. Tarel, J.-P. et al. Vision enhancement in homogeneous and heterogeneous fog. \JournalTitleIEEE Intelligent Transportation Systems Magazine 4, 6–20 (2012). Http://perso.lcpc.fr/tarel.jean-philippe/publis/itsm12.html.
  53. Zhang, C. et al. On estimating air pollution from photos using convolutional neural network. In Proceedings of the 24th ACM international conference on Multimedia (ACM, New York, NY, USA, 2016).
  54. Catboost: Unbiased boosting with categorical features. In Proceedings of the 32nd International Conference on Neural Information Processing Systems, NIPS’18, 6639–6649 (Curran Associates Inc., Red Hook, NY, USA, 2018).
  55. VisionAir. https://vision-air.github.io/. Accessed: 2023-9-17.
  56. Zhao, P. et al. Overview of MediaEval 2020: insights for wellbeing task - multimodal personal health lifelog data analysis. In MediaEval Benchmarking Initiative for Multimedia Evaluation, CEUR Workshop Proceedings (CEUR Workshop Proceedings, 2020).
  57. MNR- HCM data: a personal lifelog and surrounding environment dataset in Ho-Chi- minh city, vietnam. In ICMR-ICDAR 2020: Proceedings on Intelligent Cross-Data Analysis and Retrieval Workshop, 21–26 (2020).
  58. Comparing vision transformers and convolutional nets for safety critical systems. In SafeAI@ AAAI (2022).
  59. Full transformer framework for robust point cloud registration with deep information interaction (2021). URL https://arxiv.org/abs/2112.09385. DOI 10.48550/ARXIV.2112.09385.
Citations (3)
View on Semantic Scholar

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now