Title: A semi-automated system for smart harvesting of tea leaves

Authors: Manesh Murthi; Senthil Kumar Thangavel

Addresses: Robert Bosch Engineering and Business Solutions Private Limited, Coimbatore, India ' Department of Computer Science and Engineering, Amrita School of Engineering, Coimbatore, Amrita Vishwa Vidyapeetham, India

Abstract: Tea leaf cultivation is a major part of livelihood in hill station like Nilgiris. The conventional method of tea leaf plucking is done manually with a knife. Harvesting machines have also been designed that could quickly. This gives better result in manpower who has better experience and knowledge about terrains. The paper has proposed a semi-automatic working model that has an arm that can move around and pluck the leaves. A complete pre-processing phase has been done using keyframe extraction, rice counting, optical flow with noise model by the author in an earlier paper. This process is improved by using active contour with optical flow algorithm that minimises the region on which the tea leaf detection algorithm is applied. The second phase of the paper also suggests how deep learning approach can also be used for improving the performance of the proposed work. The proposed work is novel because it has capabilities of considering motion with keyframe capabilities and the noise model using deep learning. The proposed work has experimented with parameters like precision, recall, FAR, FRR to evaluate the nature of misclassifications.

Keywords: video analytics; noise model keyframe; Raspberry Pi; Arduino due; optical flow; rice counting; segmentation.

DOI: 10.1504/IJCAET.2020.108109

International Journal of Computer Aided Engineering and Technology, 2020 Vol.13 No.1/2, pp.125 - 153

Received: 28 Aug 2017
Accepted: 23 Oct 2017
Published online: 03 Jul 2020 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article