Fastener counting method with an improved Blendmask
Kexin Qi, Qing Xin Zhu, Xianen Zhou, Yaonan Wang, Jia Ming Feng · 2022 13th Asian Control Conference (ASCC) · 2022
The fastenings market is large. We design an automatic visual counting system, which obtains the clear image of fasteners through the combination of shaker and visual system. Datasets including in sparse and dense conditions are constructed. An improved Blendmask instance segmentation network is proposed to realize fastener counting. And we propose an improved Blendmask instance segmentation network to realize fastener counting. The experimental test is carried out by using the constructed data set. Compared to many other state-of-the-art, the proposed method can get brilliant estimations.