Ground penetrating radar (GPR), as a newly nondestructive testing technology (NDT), has been adopted to explore the spatial position and the structure of the tree roots. Due to the complexity of soil distribution and the randomness of the root position in the natural environment, it is difficult to locate the root in the GPR B-Scan image. In this study, a novel method for root detection in the B-scan image by considering both multidirectional features and symmetry of hyperbola was proposed. Firstly, a mixed dataset B-Scan images were employed to train Faster RCNN (Regions with CNN features) to obtain the potential hyperbola region. Then, the peak area and its connected region were filtered from the four directional gradient graphs in the proposed region. The symmetry test was applied to segment the intersecting hyperbolas. Finally, two rounds of coordinate transformation and line detection based on Hough transform were employed for the hyperbola recognition and root radius and position estimation. To validate the effectiveness of this approach for tree root detection, a mixed dataset was made, including synthetic data from gprMax as well as field data collected from 35 ancient tree roots and fresh grapevine controlled experiments. From the results of hyperbola recognition as well as the estimation for the radius and position of the root, our method shows a significant effect in root detection.

The outer contours of living trees are often considered as a standard circle during non-destructive testing (NDT) of internal defects using ground-penetrating radar (GPR). However, the detection of classical cross-sections (circular) lacks consideration of irregular contours, making it difficult to accurately locate the radar image of the target. In this paper, we propose a method based on the image affine transformation and the Riemann mapping principle to analyze the effect of irregular detection routes on the geometric characteristics of target reflection hyperbola. First, for the similar output phenomenon in the “hyperbola fitting”, geometric analysis and numerical simulation were performed. Then, the conversion of irregular trunk radar images and physical domain radar images was implemented using the method of image affine transformation and the Riemann mapping principle. Finally, the influence of irregular detection routes on the geometry of the target reflection curve was investigated in detail through numerical simulations and actual experiments. The numerical simulation and measurement results demonstrated that the method in this study could better reflect the imaging characteristics of the target reflection hyperbola under the irregular detection pattern. This method provides assistance to further study the defects of irregular living trees and prevents the misjudgment of targets as a result of hyperbolic distortion, resulting in a greater prospect of application.