I am designing an autonomous visual navigation system in an agriculture robot. From the farm, I got some images of the lane and now I am unable to get proper from those images. Until now, I tried many methods of image processing to detect the lanes but nothing works. I also used holistically-nested edge detection (HED) on the images but did not get the output as desired. I need suggestions that how I can detect the lanes from these images.
Currently, the field is in the Seed Sowing stage, so it is more difficult to differentiate the lanes in the field. I am attaching a sample of the lane, please give me some suggestions.
The first image is the original image, the second image is the edited one to show the lanes in the field, and the third one is formed after the canny edge detection.
Thank you
Also check https://www.researchgate.net/publication/334571033_Lane_Detection_Algorithm_for_Intelligent_Vehicles_in_Complex_Road_Conditions_and_Dynamic_Environments
Thank you Chinaza Godswill Awuchi for your valuable suggestions. I will read these papers
Kindly read this sir.
https://www.nrec.ri.cmu.edu/solutions/agriculture/other-agriculture-projects/human-detection-and-tracking.html
Thank you Jehan-Antoine Vayssade for explaining me in detail. your comments help me to understand the problem I am facing. Please give me advice, what method can I implement in the robot to detect the lanes when there is no weed and vegetation in the land. The only differentiable part is the designing of the rows. If you look closely at the image1, you can see that the vegetation lane is slightly upward than others. Is there any way to detect this difference using computer vision?
I look forward to your kind response. Thank you
- Badges
- Science topic
- Similar topics
- Geoscience
- Cartography
- Mapping
- Maps