Omnidirectional - Wheel Conveyor Path Planning and Sorting System Using Vision Algorithms

Nowadays, smart sorting and path planning systems are important in the manufacturing industry. This paper proposed an intelligent sorting and transportation system compromising a novel omnidirectional - wheel conveyor design and adaptive vision algorithm. The proposed system integrates and connects four distinct small hexagonal modules. Three omnidirectional wheels on each module enable objects to be transported freely in any direction around 360 degrees. In addition, QR code recognition enhances the system's sorting capabilities. It classifies, tracks objects, and adjusts their direction by compensating for object deviation. Experimental results showed that the module with a direct view of the camera is not affected by external light interference and achieves the highest accuracy (100<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">%</sup>) in all distance settings. The recognition accuracy gradually decreased to an acceptable level at modular around (nearly 70%). The best detection ability for a QR code was pasted at 550 mm and 620 mm from the camera. The system achieves stability during object movement at speeds of 0.15 m/s. The maximum angular deviation from theory when transporting through a single module was 5 degrees. This deviation angle also does not change much at different speeds when the object was transferred. The minimum deviation angle is negligible (less than one degree).