植被(病理学)
采样(信号处理)
增强植被指数
地理
遥感
环境科学
归一化差异植被指数
林业
植被指数
生态学
计算机科学
叶面积指数
生物
医学
滤波器(信号处理)
病理
计算机视觉
标识
DOI:10.1016/j.ufug.2021.127310
摘要
Park greenery contributes substantial health benefits for local residents. Few data source however is available to quantitatively measure the vegetation greenery along park trails. In this study, a typical park region in Homestead, Florida was selected as the study area. The iPhone panoramic images in 400 randomly generated sampling sites were examined to depict the vegetation greenery along park trails. A green vegetation extraction index (GVEI) was proposed to distinguish the green vegetation in the images, and the green vegetation index was computed to grade the sampling sites based on the GVEI results. The results demonstrated that the overall (kappa coefficient) accuracies of green vegetation classification in the sampling sites using the GVEI all exceeded 0.90 (0.87). The GVEI values obtained from iPhone and commonly used Google Street View (GSV) images showed a high correlation coefficient of 0.94, which indicated that the iPhone panoramic images were compatible with the GSV images, and suited for green vegetation identification and monitoring. The green view index of sampling sites showed discrepant patterns, 50 sites possessed the green view index of more than 30 %, indicating that people in these areas can enjoy visual and spiritual pleasure. Park managers and planners may strive to improve the vegetation greenery in the other sites to offer better health benefits for people. The method established in this study can be used as a reference for vegetation greenery monitoring in other parks.
科研通智能强力驱动
Strongly Powered by AbleSci AI