Health Assessment of Eucalyptus Trees Using Siamese Network from Google Street and Ground Truth Images

Urban greenery is an essential characteristic of the urban ecosystem, which offers various advantages, such as improved air quality, human health facilities, storm-water run-off control, carbon reduction, and increase in property values. Therefore, identification continuous monitoring vegetation (trees) vital importance for our lifestyle. This paper proposes a deep learning-based network, Siamese convolutional neural network (SCNN), combined with modified brute-force-based line-of-bearing (LOB) algorithm that evaluates Eucalyptus trees healthy or unhealthy identifies their geolocation real time from Google Street View (GSV) ground truth images. Our dataset represents trees’ details multiple viewpoints, scales different shapes to texture. The experiments were carried out Wyndham city council area state Victoria, Australia. approach obtained average accuracy 93.2% identifying after training on around 4500 images testing 500 study helps tree issues dead automated way can facilitate green management assist local make decisions about plantation improvements looking trees. Overall, this shows even complex background, most be detected by learning time.