This thesis presents the design, implementation, and experimental evaluation of two software libraries for vision-guided robotic picking in industrial feeding machine applications, developed during an internship at EPF Elettrotecnica S.r.l. and integrated into the company’s Supata® flexible feeding platform. The first library addresses the problem of adaptive stability detection in vibrating feeding environments. Rather than relying on a fixed timer to determine when objects on the vibrating table have come to rest, the library continuously monitors the scene by analysing inter-frame pixel variation and declares stability only when a configurable criterion has been satisfied for a sufficient number of consecutive frames.