СЕГМЕНТАЦИЯ ИЗОБРАЖЕНИЙ С ИСПОЛЬЗОВАНИЕМ ИИ ДЛЯ КОЛИЧЕСТВЕННОГО АНАЛИЗА СТЕАРАТА ТРИЭТАНОЛАМИНА КАК ПОВЕРХНОСТНО-АКТИВНОГО ВЕЩЕСТВА ПРИ ЛИКВИДАЦИИ НЕФТЯНЫХ РАЗЛИВОВ
Стр.38-47
DOI: 10.32758/2782–3040-2025-0-4-38-47
УДК 620.193
Z.Z. Aghamaliyev, E.A. Aydinsoy, D.B. Aghamaliyeva, K.A. Mammadova
(Y.H.Mamedaliyev’s Institute of Petrochemical Processes of the Ministry of Science and Education, Baky)
Abstract. Efficient surfactants are vital for oil spill mitigation and environmental protection. This study evaluates a 5 % solution of triethanolamine stearate in ethyl alcohol, a newly synthesised surfactant, for its oil-collecting efficiency using an AI-powered image segmentation model. By applying the surfactant to oil-contaminated water samples and analysing high-resolution images, we quantified oil recovery efficiency, with the model achieving a mean average precision (mAP) of 99.5 %. Proportionality constants (K) were calculated, showing the highest K-value of 77.44 in distilled water with Oil Rocks petroleum, recovering up to 88.6% of the oil. In more challenging water conditions, Neftchala and Pirallahi waters yielded K-values of 36.83, with recovery rates of 83.5 %, while Sumgayit water recorded the lowest K-values of 11.46, achieving 70.5 % recovery. These findings highlight the solution’s promising oil-collecting capacity across varied environments, suggesting its potential for scalable, effective oil spill response.
Keywords: Surface Active Agents, Oil Spill Mitigation, Image Segmentation, Environmental Cleanup, Marine Oil Spill Remediation.