Smart Scarecrow Systems for Crop Protection: A Survey

Birds and wild or stray animals remain a persistent cause of pre-harvest crop loss, and the cheapest defences against them lose force over time. Scarecrows, reflective tape, noise makers and fences are inexpensive, yet animals learn that a stimulus which never changes carries no threat. This habituation is the central weakness of passive deterrence and the reason researchers have turned to systems that sense an intruder and respond only when one is present [1]–[4].Over the last decade, "smart scarecrow" designs have multiplied. They range from low-cost devices that pair a passive infrared (PIR) sensor with an Arduino-class microcontroller, a buzzer and a servo, to camera-based platforms that run deep-learning detectors on a Raspberry Pi or an embedded neural accelerator [5], [6]. Most reported work is spread across student theses, regional journals and conference papers, and the designs are rarely compared on common terms. Small farms, orchards and nurseries need a clear view of what is known, what is proven, and what is still guesswork before they invest in such a system.