• S. Duari
• K.Rajkumar

The mining conveyor belt is a crucial component of the coal mine belt conveyor system, responsible for transporting materials and transmitting power. Due to the harsh working conditions, these belts are prone to damage. To address this issue, we reclassified and defined various types of conveyor belt damage and created a specialized dataset for this purpose. We then proposed a novel detection method capable of simultaneously identifying multiple faults using an improved YOLOv3 algorithm. In this enhanced algorithm, Efficient Net replaces Darknet53 as the backbone feature extraction network, achieving a balance between network depth, width, and image resolution to enhance accuracy within limited computing resources. Experimental results demonstrate that our improved algorithm excels in both detection speed and accuracy, achieving a speed of 42 FPS and a mean average precision of 97.26%. Compared to the original YOLOv3 algorithm, our method increases accuracy by 10.4% and speed by 45.9%, offering innovative solutions for ensuring the safe and stable operation of conveyor belts.

Belt conveyor, Conveyor belt, Deep learning, Machine vision, Damage detection.

"Deep Learning-Based Damage Detection of Mining Conveyor Belts", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.11, Issue 5, page no.n758-n767, May-2024, Available :http://www.jetir.org/papers/JETIR2405D98.pdf

"Deep Learning-Based Damage Detection of Mining Conveyor Belts", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.11, Issue 5, page no. ppn758-n767, May-2024, Available at : http://www.jetir.org/papers/JETIR2405D98.pdf