Experience of implementation and approaches for further automation in unmanned aerial counting of pacific salmon in the Okhotsk district of Khabarovsk Region

Abstract

Unmanned aerial counting of pacific salmon in the rivers of the Okhotsk district (Khabarovsk Region) in 2021–2023 is analyzed. Estimates of the spawning redds number and the spawners abundance are presented for the monitoring sites. Prospects for using these indices for forecasting the stocks are discussed. Specifics of unmanned counting are considered, and options for further development of the methodology are proposed. The method of unmanned counting is suitable for information support of the stock forecasts only, but can be modified for application in fishery management and spawning grounds inventory. Possibilities for further automation in the unmanned counting of pacific salmon are considered on the base of the authors’ experience and literature information. Proposed approaches to automation are outlined for key stages of unmanned counting, as spatial coverage planning, aerial photography acquisition, photogrammetric processing, vectorization, and abundance estimation. Features of the automation implementation depend on purposes of the counting and are different for the stock forecasting, fishery management, and spawning grounds inventory. Adequate automation should be based on ensuring consistency and mutual balance of approaches used at different stages of the counting.

Keywords

• pacific salmon
• unmanned aerial vehicle (uav)
• photogrammetry
• geoinformation system (gis)
• vectorization
• automation
• artificial intelligence