Treffer: SUSTAV AUTOMATIZIRANE INSPEKCIJE BOCA U PROIZVODNOM PROCESU UZ PRIMJENU RASPBERRY PI RAČUNALNE PLATFORME ; AUTOMATED BOTTLE INSPECTION SYSTEM IN THE PRODUCTION PROCESS USING RASPBERRY PI COMPUTING PLATFORM
University of Split. University Department of Professional Studies. Department of Electrical Engineering.
Weitere Informationen
Automatizacija nadzora ispravnosti proizvoda u stvarnom vremenu postaje sve važnija u proizvodnim procesima. U ovom radu prikazana je izrada sustava za detekciju i analizu boca pomoću različitih senzora, računalnog vida i mikroračunala Raspberry Pi 3B+. Sustav se sastoji od senzora mase, senzora ambijentalnog osvjetljenja, kamere te Python programskog koda, a uz pomoć biblioteke OpenCV i ostalih. Korištenjem senzorskih podataka i slike, sustav detektira prisutnost boce, ispituje razinu osvjetljenja, prepoznaje etiketu i čep, određuje boju i razinu tekućine u boci te uspoređuje masu s definiranim pragovima. U slučaju da neki od kriterija nisu zadovoljeni, boca se klasificira kao neispravna. Svi podaci i slike rezultata šalju se i pohranjuju u realnom vremenu u Firebase bazu podataka. Sustav predstavlja senzorsku mrežu povezanu sa bazom podataka u stvarnom vremenu, što omogućuje brzu, pouzdanu i temeljitu procjenu kvalitete proizvoda. ; Automation of real-time product monitoring is becoming increasingly important in manufacturing processes. This scientific writing demonstrates the development of a system for detecting and analyzing bottles using various sensors, computer vision, and the Raspberry Pi 3B+ microcomputer. The system consists of a mass sensor, an ambient light sensor, a camera, and Python programming code, with assistance of the OpenCV library and others. By using sensor data and image, the system detects presence of a bottle, examines the light level, recognizes the label and cap, determines the color and liquid level in the bottle, and compares mass with defined thresholds. In case some of criteria are not met, bottles are classified as defective. All data and images of the results are sent and stored in real time in the Firebase database. The system represents a sensor network connected to a real-time database, which enables fast, reliable, and thorough examination of product quality.