Spectral-analysis-based extraction of land disturbances arising from oil and gas development in diverse landscapes

摘要: Accurate and frequent monitoring of land surface changes arising from oil gas exploration extraction is a key requirement for the responsible sustainable development these resources. Petroleum deposits typically extend over large geographic regions but much infrastructure required recovery takes form numerous small-scale features (e.g., well sites, access roads, etc.) scattered landscape. Increasing exploitation will increase presence disturbances in heavily populated regions. An object-based approach proposed to utilize RapidEye satellite imagery delineate sites related roads diverse complex landscapes, where also arise other human activities, such as forest logging agriculture. A simplified change vector approach, adaptable operational use, introduced identify on based red–green spectral response spatial attributes candidate object size proximity roads. Testing techniques has been undertaken with multitemporal two test located at Alberta, Canada: one was predominant natural landscape dominated by intensive agricultural activities. Accuracies 84% 73%, respectively, have achieved identification site road fully automated processing. Limited manual relabeling selected image segments can improve accuracies 95%.