Scholarlink

To solve the current quality detection difficulty that the detection of nutrient content for sea cucumbers is assessed using destructive biochemical methods without considering non-destructive detection of nutrients such as saponins, this paper proposes an ensemble learning model based on near-infrared spectroscopy for the non-destructive detection of saponins and other important nutrients in intravital sea cucumbers. The model establishes a quantitative relationship between the contents of nutrition in sea cucumbers and their spectral characteristics, such as saponins, fats, and proteins. Experimental results show that the proposed model can effectively detect the content of saponins, fats, and proteins in sea cucumbers, with coefficients of determination (R2) of 0.92, 0.96, and 0.98, and root mean square errors (RMSE) of 0.63, 0.48, and 0.41, respectively. This proposed approach realizes the non-destructive detection of the nutritional quality for sea cucumbers, especially for saponins, which overcomes the drawbacks of traditional detection methods that are time-consuming, labor-intensive, expensive, and destructs the tested samples.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Non-destructive Detection of Sea Cucumber Saponins and Other Nutrients Based on Hyperspectral Imaging and Ensemble Learning

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Yunhao Zhang,
  • Juan Li,
  • Peng Lou,
  • Hao Yang,
  • Yuefeng Zhang,
  • Xiaodong He

摘要

To solve the current quality detection difficulty that the detection of nutrient content for sea cucumbers is assessed using destructive biochemical methods without considering non-destructive detection of nutrients such as saponins, this paper proposes an ensemble learning model based on near-infrared spectroscopy for the non-destructive detection of saponins and other important nutrients in intravital sea cucumbers. The model establishes a quantitative relationship between the contents of nutrition in sea cucumbers and their spectral characteristics, such as saponins, fats, and proteins. Experimental results show that the proposed model can effectively detect the content of saponins, fats, and proteins in sea cucumbers, with coefficients of determination (R2) of 0.92, 0.96, and 0.98, and root mean square errors (RMSE) of 0.63, 0.48, and 0.41, respectively. This proposed approach realizes the non-destructive detection of the nutritional quality for sea cucumbers, especially for saponins, which overcomes the drawbacks of traditional detection methods that are time-consuming, labor-intensive, expensive, and destructs the tested samples.