Improvements in accuracy of single camera terrain classification

Abbas, Syed Muhammad; Muhammad, Abubakr; Mehdi, Syed Atif; Berns, Karsten

doi:10.1109/icar.2013.6766493

Cited by 6 publications

(3 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Techniques exist for identifying classes of traversable terrains in the form of multi-class classification approaches [12]- [18]. While these approaches could also enable more precise navigation decisions the problem is that identifying that terrain is of one class or another does not, considered alone, provide any information about the relative traversability of the terrain.…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Terrain Roughness Identification for High-Speed UGVs

Wilson¹,

Ramírez-Serrano²

2014

JACR

View full text Add to dashboard Cite

show abstract

Section: Discussionmentioning

confidence: 99%

“…Visual data from a camera image has been used to train machine learning techniques such as SVMs, Extreme Learning Machines (ELM), NNs, and Bayesian classifiers [12]- [18]. Work by Abbas et al used colour and texture features along with a SVM to classify six terrain types with up to 97% accuracy.…”

Section: Related Workmentioning

confidence: 99%

Terrain Roughness Identification for High-Speed UGVs

Wilson¹,

Ramírez-Serrano²

2014

JACR

View full text Add to dashboard Cite

show abstract

“…An accelerometer was used for a land vehicle [2] while an Inertial Measurement Unit (IMU) was used for a legged robot [3] to extract vibration-based features for the road terrain classifications. Camera was also applied to obtain the colour and textured features for outdoor road terrain identification [4]. The LRF's measurement was also employed with an IMU to improve the classification performance [5].…”

Section: Background and Related Workmentioning

confidence: 99%

Two-Stage Road Terrain Identification Approach for Land Vehicles Using Feature-Based and Markov Random Field Algorithm

et al. 2018

View full text Add to dashboard Cite

Road terrain identification is one of the important tasks for driving assistant system or autonomous land vehicle. It plays a key role to improve driving strategy and enhance the fuel efficiency. In this paper, a two-stage approach using multiple sensors is presented. In the first stage, a feature-based identification approach is performed using single-sensor: an accelerometer, a camera, a downward-looking and a forward-looking laser range finders (LRFs), respectively. This produces four classification label sequences. In the second stage, a Majority-Vote is implemented for each label sequences to match them into synchronized road patches. Then a Markov Random Field (MRF) model is designed to generate the final optimized identification results to improve the forward-looking LRF. This approach enables the vehicle to observe the upcoming road terrain before moving onto it by fuses all the classification results using MRF algorithm. The experiments show this approach improved the terrain identification accuracy and robustness significantly for some familiar road terrains.

show abstract