|
CSME 2019/12
Volume 40 No.6
:
683-692
|
|
Study on Navigation Line Detection in L.barbarum Garden Based on Genetic Algorithm
Mao-Qiang Lia and Zhi-Feng Liua
aSchool of Mechanical Engineering, Hefei University of Technology, Hefei, Anhui 230009, China.
|
Abstract:
The form for a “general self-moving host platform + operation module” must be established to realize intelligent production in the Lycium barbarum (L.barbarum). Region segmentation between the plant column and the soil and detection of the navigation line are two challenges to realizing the autonomous navigation of self-moving platforms. For region segmentation between the plant column and the soil, a segmentation method that based on neural networks and Otsu’s method is proposed. Experiments are performed that demonstrate effective image segmentation using this approach. A noise removal technique is proposed that can effectively remove the noise from the plant column region and the noise from the soil region. A parameterized octagonal template is constructed for the navigation line detection problem after image segmentation, which is matched to the segmentation image. The matching overlap is defined as the fitness function, and the octagonal template parameter is optimized based on a genetic algorithm. Then, the octagon midline is extracted from the optimized octagonal template as the navigation line. The experiments demonstrate that the method can effectively detect the navigation line using this method, which lays a foundation for the precise navigation of self-moving platforms in complex and unstructured L.barbarum Garden environments.
|
|
Keywords: Lycium barbarum garden, Self-moving platform, image segmentation, navigation line detection.
|
Download PDF
|
|
*Corresponding author; e-mail:
|
©
2019
CSME , ISSN 0257-9731
|