Intelligent Fruit Picking Robot based on Deep Learning

Zhenyun Sun ； Baofeng Ji ； Xianxian Shi ； Hui Han ； Bingyi Dong ； Mi Zhang ； Yunchao Zhang ； Tongfei Li

《International Core Journal of Engineering》 8卷6期 (2022/06) Pp. 70-73

https://doi.org/10.6919/ICJE.202206_8(6).0010

摘要

Aiming at orchard environment, an intelligent picking robot based on deep learning was proposed. This picking robot adopts the convolutional neural network model based on the YOLOV5 framework, and trains the model through advanced deep learning technology, so that the robot can recognize the fruit through computer vision, and realizes the stereo positioning of the fruit by using the principle of binocular vision. In this paper, the method of fruit target detection and recognition is designed, so that the robot system can achieve the three-dimensional positioning and precise picking of target fruit.

關鍵字

Deep Learning ； Convolutional Neural Network ； Computer Vision ； Target Recognition ； Target Location

參考文獻

Fu, L., Tola, E., Al-Mallahi, A., Li, R., and Cui, Y. (2019). A novel image processing algorithm to separate linearly clustered kiwifruits. Biosyst. Eng. 183, 184–195.

Fu, L., Tola, E., Al-Mallahi, A., Li, R., and Cui, Y. (2019). A novel image processing algorithm to separate linearly clustered kiwifruits. Biosyst. Eng. 183, 184–195.

Gong, J. Y., Jiang, B., and Shen, Q. K. (2019). Adaptive fault-tolerant neural control for large-scale systems with actuator faults. Int. J. Control Automat. Syst. 17, 1421–1431. doi: 10.1109/TNNLS. 2016. 2598580.

Horea, M., and Mihai, O. (2018). Fruit recognition from images using deep learning. Acta Univ. Sapientiae Inform. 10, 26–42. doi: 10.3389/fpls.2019.00611.

Esteva A, Chou K, Yeung S, et al. Deep learning-enabled medical computer vision[J]. NPJ digital medicine, 2021, 4(1): 1-9.