Teknik identifikasi saluran irigasi pada citra satelit resolusi tinggi dengan penggabungan komposit RGB, indeks saluran, dan interpretasi visual

Abstract

One of the important programs to support the national food security is the development and rehabilitation of the irrigation network infrastructure. Spatial data on the location of existing irrigation networks becomes very important information for the policy of regional expansion of irrigation and monitoring of the damage of irrigation network infrastructure. High spatial resolution satellite imagery such as SPOT 7 and the Pleiades is capable of visually exposing the earth's surface objects, such as roads, rivers, and irrigation canals. Research was conducted to develop an irrigation canal identification procedure using high-resolution imagery. The identification procedure is done by combining the RGB composite image with false color, the canal index, and visual interpretation by recognizing the characteristics of the irrigation canal. Development of RGB composites and canal index was conducted based on the spectral pattern differences from the irrigation canal with spectral patterns of some similar objects. RGB composite imagery combined with canal index can clarify and isolate irrigation canals from surrounding object with some constraints in the canal width and crop shadows. Imagery interpretation techniques can ensure precise identification and distinguish irrigation canals with crop shadows. An image of SPOT 6/7 can be utilized for canal identification of more than 4 m in width, and the image of Pleaides for canal of less than 4 m in width. The total accuracy of the irrigation canal identification results using Pleaides imagery was around 82%.