Technological Development to Specifically Extract Shoots from Images of Tea Canopy Surfaces and Estimate Numbers of Opened Leaves

DOI Web Site 1 References
  • Hasegawa Kazuya
    Tea Research Center, Shizuoka Prefectural Research Institute of Agriculture and Forestry
  • Furuya So
    Tea Research Center, Shizuoka Prefectural Research Institute of Agriculture and Forestry
  • Suzuki Toshikazu
    Tea Research Center, Shizuoka Prefectural Research Institute of Agriculture and Forestry
  • Oishi Tetsuya
    Tea Research Center, Shizuoka Prefectural Research Institute of Agriculture and Forestry

Bibliographic Information

Other Title
  • 茶樹冠面画像から新芽を特異的に抽出し開葉数を推定する技術の開発

Description

<p>We developed a simple method to estimate the growth stage of tea shoots to reduce the labor involved of tea farmers. We developed a program using Python to estimate the number of opened leaves of tea shoots by capturing pictures of the surface of tea trees using a smartphone and mask processing, with the exception of new shoots. The accuracy of the mask processing was 84%, and the process was successful. The saturation values of the extracted shoots were correlated with the number of opened leaves regardless of the camera used, and a model was developed to estimate the number of open leaves based on the saturation values. Using this model, we compared the estimated number of opened leaves with the actual number of opened leaves measured at eight tea gardens in Shizuoka Prefecture; the root mean square error of prediction (RMSEP) value of the estimation was 0.33 leaves.</p>

Journal

References(1)*help

See more

Keywords

Details 詳細情報について

Report a problem

Back to top