Coffee has been cultivated as the secondary produce for decades in Girimarto, Wonogiri, however, the selective picking practice remain alient for local farmers. Selective picking is considered impractical due to time consumption and laborous work that farmers should carry out. The community service project designs and implements image recognition technology to help acquire coffee-cheery ripeness condition. Adopting a geolocation, the appropriate routing strategies, would enable farmers to selectively pick the red cherries in a systematic sequence. The image processing technology was applied by adopting Raspberry Pi microcomputer, Raspberry Pi Camera Board, version 2, and OpenCV programming language. The transition to selective picking and the subsequent post-harvest technology would likely produce high-quality green bean coffee. It is expected that the income of smallholder coffee farmers will gradually be increasing.