AUTHOR=Yang Zhenling , Yang Yang , Li Chaorong , Zhou Yang , Zhang Xiaoshuang , Yu Yang , Liu Dan 

TITLE=Tasseled Crop Rows Detection Based on Micro-Region of Interest and Logarithmic Transformation

JOURNAL=Frontiers in Plant Science

VOLUME=Volume 13 - 2022

YEAR=2022

URL=https://www.frontiersin.org/journals/plant-science/articles/10.3389/fpls.2022.916474

DOI=10.3389/fpls.2022.916474

ISSN=1664-462X

ABSTRACT=While deep learning methods have made great progress in agricultural machinery navigation, training a deep neural network usually requires a lot of computing resources. Therefore, how to accurately detect the tasseled crop rows for agricultural machinery navigation with a simple and energy-efficient method remains an open question. In this paper, we propose a new crop rows detection method at the tasseling stage of maize fields for agrarian machinery navigation. The whole work is achieved mainly through image enhancement and feature point selection by micro-region of interest (ROI). In the proposed method, we first augment the distinction between the tassels and background by the logarithmic transformation in RGB space, and then the image is transformed to hue-saturation-value (HSV) space to extract the tassels. Secondly, the ROI is roughly selected and updated using the bounding box until the multiple-region of interest (multi-ROI) is determined. We further propose a Micro-ROIs-based method to extract the feature points and fit them to determine the crop rows detection lines. Finally, the bisector of the acute angle formed by the two detection lines is used as the field navigation line. The experimental results show that the algorithm has good robustness and can accurately identify crop rows. The average computation time of this algorithm for processing a single frame is 312.3ms, the average error angle of the navigation line is 1.49°, and the accuracy is more than 98%, which can meet the real-time and accuracy requirements of agricultural vehicles' navigation in maize fields.