| dc.contributor.author | Павлик, О. В. | uk |
| dc.contributor.author | Романюк, О. Н. | uk |
| dc.contributor.author | Pavlyk, O. V. | en |
| dc.contributor.author | Romanyuk, O. N. | en |
| dc.date.accessioned | 2026-09-03T10:55:40Z | |
| dc.date.available | 2026-09-03T10:55:40Z | |
| dc.date.issued | 2026 | |
| dc.identifier.citation | Павлик О. В., Романюк О. Н. Verification methods for cleanup results in software tools for environmental monitoring // Матеріали Міжнародної науково-практичної інтернет-конференції «Молодь в науці: дослідження, проблеми, перспективи (МН-2026)», м. Вінниця, 22-26 червня 2026 р. Електрон. текст. дані. 2026. URI: https://conferences.vntu.edu.ua/index.php/mn/mn2026/paper/view/29816. | uk |
| dc.identifier.uri | https://ir.lib.vntu.edu.ua/handle/123456789/53892 | |
| dc.description.abstract | The paper examines verification methods for cleanup results in environmental monitoring software tools. Approaches to automated validation of photo evidence using computer vision, geolocation confirmation, crowdsourced validation, and blockchain registration are analyzed. The advantages and limitations of each method are identified. A comprehensive architecture for a verification subsystem of a mobile application for organizing urban cleanup events is proposed. | en |
| dc.description.abstract | У тезах розглядаються методи верифікації результатів прибирань у програмних засобах екологічного моніторингу. Проаналізовано підходи до автоматизованої перевірки фотодоказів із використанням комп’ютерного зору, геолокаційного підтвердження, краудсорсингової валідації та блокчейн-реєстрації. | uk |
| dc.language.iso | en_US | en_US |
| dc.publisher | ВНТУ | uk |
| dc.relation.ispartof | Матеріали Міжнародної науково-практичної інтернет-конференції «Молодь в науці: дослідження, проблеми, перспективи (МН-2026)», м. Вінниця, 22-26 червня 2026 р. | uk |
| dc.relation.uri | https://conferences.vntu.edu.ua/index.php/mn/mn2026/paper/view/29816 | |
| dc.subject | верифікація | uk |
| dc.subject | екологічний моніторинг | uk |
| dc.subject | комп’ютерний зір | uk |
| dc.subject | геолокація | uk |
| dc.subject | краудсорсинг | uk |
| dc.subject | блокчейн | uk |
| dc.subject | мобільний застосунок | uk |
| dc.subject | прибирання міста | uk |
| dc.subject | verification | en |
| dc.subject | environmental monitoring | en |
| dc.subject | computer vision | en |
| dc.subject | geolocation | en |
| dc.subject | crowdsourcing | en |
| dc.subject | blockchain | en |
| dc.subject | mobileapplication | en |
| dc.subject | urban cleanup | en |
| dc.title | Verification methods for cleanup results in software tools for environmental monitoring | en |
| dc.type | Thesis | |
| dc.identifier.udc | 004.9:502.17 | |
| dc.relation.references | the GPS coordinates embedded in submitted media with the registered cleanup event location. A spatial tolerance parameter d (measured in metres) defines the acceptable radius: if the Euclidean distance between the reported point P_r and the event centroid P_e satisfies d(P_r, P_e) <= d_max, the submission is considered spatially valid [5]. Temporal checks further require that the submission timestamp falls within the active event window. Modern platforms additionally use sensor fusion – combining GPS, Wi-Fi triangulation, and accelerometer data – to improve location accuracy and detect stationary spoofing attempts. Crowdsourced and Peer Validation Peer-review mechanisms leverage the community itself as a verification layer. In this model, a randomly selected subset of registered participants reviews submitted evidence and votes on its authenticity [6]. The consensus decision is computed using weighted majority voting, where trust weights w_i are assigned to reviewers based on their historical accuracy and participation history. This approach scales well with community size but introduces latency and depends on sufficient active reviewer availability. Reputation systems and anti-collusion algorithms must be carefully designed to prevent coordinated fraud. Blockchain-Based Registration Blockchain technology provides immutable, timestamped records of cleanup events and verification decisions [7]. Each verified submission is encoded as a transaction containing a cryptographic hash of the evidence media, event metadata, and participant identifier. The distributed ledger structure prevents post-hoc modification and enables transparent audit trails. Smart contracts can automate reward distribution upon successful verification. Despite these advantages, on-chain storage of large media files is impractical; therefore, hybrid architectures using IPFS (InterPlanetary File System) for off-chain storage with on-chain hash anchoring are preferred. Proposed Integrated Architecture Based on the analysis above, a layered verification pipeline is proposed for the target mobile application. The first layer performs automated EXIF metadata parsing and GPS boundary checking. The second layer applies a lightweight on-device CNN model for preliminary litter detection. Inconclusive cases are escalated to the third layer – crowdsourced peer review. Finalized results are anchored to a permissioned blockchain ledger for audit purposes. This multi-stage approach balances automation efficiency with verification reliability and minimizes the load on human reviewers [8]. Conclusions The paper has systematically reviewed four major categories of verification methods for urban cleanup results: automated media analysis, geolocation validation, crowdsourced peer review, and blockchain-based registration. Each method presents distinct trade-offs between automation degree, fraud resistance, and implementation complexity. The proposed integrated multi-layer architecture combines the complementary strengths of these approaches and provides a practical foundation for the verification subsystem of the target software platform for organizing community ecological events in urban environments. REFERENCES | en |
| dc.relation.references | Gharaibeh A. et al. Smart cities: A survey on data management, security, and enabling technologies. IEEE Communications Surveys & Tutorials. 2017. Vol. 19. No. 4. P. 2456–2501. | en |
| dc.relation.references | See L. et al. Crowdsourcing, citizen science or volunteered geographic information? The current state of play. ISPRS International Journal of Geo-Information. 2016. Vol. 5. No. 5. P. 55. | en |
| dc.relation.references | Hamari J., Koivisto J., Sarsa H. Does gamification work? – A literature review of empirical studies on gamification. 47th Hawaii International Conference on System Sciences. 2014. P. 3025–3034. | en |
| dc.relation.references | Majchrowska S. et al. Deep learning-based waste detection in natural and urban environments. Waste Management. 2022. Vol. 138. P. 274–284. | en |
| dc.relation.references | Zandbergen P. A. Accuracy of iPhone locations: A comparison of assisted GPS, WiFi and cellular positioning. Transactions in GIS. 2009. Vol. 13. No. s1. P. 5–25. | en |
| dc.relation.references | Brabham D. C. Crowdsourcing as a model for problem solving. Convergence. 2008. Vol. 14. No. 1. P. 75–90. | en |
| dc.relation.references | Nakamoto S. Bitcoin: A Peer-to-Peer Electronic Cash System. 2008. URL: https://bitcoin.org/bitcoin.pdf (date of access: May 25, 2026). | en |
| dc.relation.references | Benet J. IPFS – Content Addressed, Versioned, P2P File System. arXiv preprint arXiv:1407.3561. 2014. | en |
