Horizon Detection Using Machine Learning Techniques

Fefilatyev, Sergiy; Smarodzinava, Volha; Hall, Lawrence O.; Goldgof, Dmitry B.

doi:10.1109/icmla.2006.25

Cited by 76 publications

(62 citation statements)

References 2 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The execution time is relatively long and not compatible with a real time functionality. Certain authors use machine learning [16] to segment the sky and the earth. However, this solution is dependent on the number of images enabling machine learning and the diversity of situations and exposures complicate this learning.…”

Section: Related Workmentioning

confidence: 99%

Real-time infinite horizon tracking with data fusion for augmented reality in a maritime operations context

2013

View full text Add to dashboard Cite

In this paper we propose a robust method for real time detection of the horizon (ie. separation line between the coast and sea) in a maritime context whose ultimate goal is to combine a video stream and electronic charts. We present the combination of an image analysis algorithm with information from an inertial unit. The initial aim is to filter out environmental conditions and noise. The second aim is to detect the visible horizon using both a classic video stream and a thermal video stream taking into consideration the nonnegligible presence of this type of camera on board vessels. We then present an assessment of the algorithm's performance with regard to the rate of detection of the horizon and the impact of the image's resolution on the calculational loop's execution time. The purpose of developing this method for detecting the visible horizon is to create an indirect vision functionality in augmented reality. We want to combine information from charts and the video stream in a single display. We use the information regarding the horizon's position in the image to divide it into different areas whose transparency enables either the video or charts to be given priority.

show abstract

Section: Related Workmentioning

confidence: 99%

Real-time infinite horizon tracking with data fusion for augmented reality in a maritime operations context

2013

View full text Add to dashboard Cite

show abstract

“…Horizon line detection has many important applications such as ship detection, flight control and port security [2]. Recently, Baatz et al [8] have demonstrated the idea of using the horizon line for visual geo-localization of images in mountainous terrain.…”

Section: Introductionmentioning

confidence: 99%

“…Their choice of decision trees is motivated by the computational efficiency achieved at run time to perform sky segmentation for static obstacle avoidance by Micro Air Vehicles (MAVs). They have extended the features used in [2] by introducing new features such as cornerness, grayness, and Fisher discriminant features. In comparison to [2], they used an extended database to train their classifier and a large number of features; hence their approach is more robust and capable of finding non-linear horizon lines.…”

Section: Introductionmentioning

confidence: 99%

“…The horizon line is then discriminated from the remaining candidates using length and continuity constraints. Fefilatyev et al [2] have proposed a machine learning approach to find the horizon line by segmenting the sky and non-sky regions assuming that the horizon line is straight. Various color and texture features (e.g., mean intensity value of three color channels, entropy, smoothness, uniformity, third moment etc.)…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A Machine Learning Approach to Horizon Line Detection Using Local Features

Ahmad

Bebis

Regentova

et al. 2013

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Abstract. Planetary rover localization is a challenging problem since no conventional methods such as GPS, structural landmarks etc. are available. Horizon line is a promising visual cue which can be exploited for estimating the rover's position and orientation during planetary missions. By matching the horizon line detected in 2D images captured by the rover with virtually generated horizon lines from 3D terrain models (e.g., Digital Elevation Maps(DEMs)), the localization problem can be solved in principle. In this paper, we propose a machine learning approach for horizon line detection using edge images and local features (i.e., SIFT). Given an image, first we apply Canny edge detection using various parameters and keep only those edges which survive over a wide range of thresholds. We refer to these edges as Maximally Stable Extremal Edges (MSEEs). Using ground truth information, we then train an Support Vector Machine (SVM) classifier to classify MSEE pixels into two categories: horizon and non-horizon. Each MSSE pixel is described using SIFT features which becomes input to the SVM classifier. Given a novel image, we use the same procedure to extract MSSEs; then, we classify each MSEE pixel as horizon or non-horizon using the SVM classifier. MSEE pixels classified as horizon are then provided to a Dynamic Programming shortest path finding algorithm which returns a consistent horizon line. In general, Dynamic Programming returns different solutions (i.e., due to gaps) when searching for the optimum horizon line in a left-to-right or right-to-left fashion. We use the actual output of the SVM classifier to resolve ambiguities in places where the left-to-right and right-to-left solutions are different. The final solution, is typically a combination of edge segments from the left-to-right or right-to-left solutions. Moreover, we use the SVM classifier to fill in small gaps in the horizon line; this is in contrast to the traditional dynamic programming approach which relies on mere interpolation. We report promising experimental results using a set of real images.

show abstract

“…There are several applications of detected horizon including smooth navigation of unmanned aerial vehicles (UAVs) [17], [14], [23] and micro air vehicles (MAVs) [12], [13], [15], augmented reality [20], visual geo-localization and annotation of mountain/desert imagery [22], port security [16], and outdoor vehicle localization. Previous horizon line detection methods can be grouped into two major categories; (i) methods modeling sky and nonsky regions using machine learning [17], [12], [13], [16], [14], [15], [21] and (ii) methods relying on edge detection [19], [5]. Some attempts [1], [2], [20] have been made to combine these two approaches by eliminating non-horizon edges using classification; however, these attempts also fall under the second category as horizon detection is effectively based on edges.…”

Section: Introductionmentioning

confidence: 99%

Fusion of edge-less and edge-based approaches for horizon line detection

Ahmad

Bebis

Nicolescu

et al. 2015

2015 6th International Conference on Information, Intelligence, Systems and Applications (IISA)

View full text Add to dashboard Cite

Abstract-Horizon line detection requires finding a boundary which segments an image into sky and non-sky regions. It has many applications including visual geo-localization and geotagging, robot navigation/localization, and ship detection and port security. Recently, two machine learning based approaches have been proposed for horizon line detection: one relying on edge classification and the other relying on pixel classification. In the edge-based approach, a classifier is used to refine the edge map by removing non-horizon edges. The refined edge map is then used to form a multi-stage graph where dynamic programming is applied to extract the horizon line. In the edge-less approach, classification is used to obtain a confidence of horizon-ness at each pixel location. The horizon line is then extracted by applying dynamic programming on the resultant dense classification map rather than on the edge map. Both approaches have shown to outperform the classical approach where dynamic programming is applied on the non-refined edge map. In this paper, we provide a comparison between the edge-less and edge-based approaches using two challenging data sets. Moreover, we propose fusing the information about the horizon-ness and edge-ness of each pixel. Our experimental results illustrate that the proposed fusion approach outperforms both the edge-based and edge-less approaches.

show abstract

Horizon Detection Using Machine Learning Techniques

Cited by 76 publications

References 2 publications

Real-time infinite horizon tracking with data fusion for augmented reality in a maritime operations context

Real-time infinite horizon tracking with data fusion for augmented reality in a maritime operations context

A Machine Learning Approach to Horizon Line Detection Using Local Features

Fusion of edge-less and edge-based approaches for horizon line detection

Contact Info

Product

Resources

About