A Comprehensive Review of Image Line Segment Detection and Description: Taxonomies, Comparisons, and Challenges

TL;DR

Based on their mechanisms, two taxonomies for line segment detection and description are presented to provide researchers with an overall picture and deep understanding and guide researchers in selecting the best method for their intended vision applications.

Abstract

An image line segment is a fundamental low-level visual feature that delineates straight, slender, and uninterrupted portions of objects and scenarios within images. Detection and description of line segments lay the basis for numerous vision tasks. Although many studies have aimed to detect and describe line segments, a comprehensive review is lacking, obstructing their progress. This study fills the gap by comprehensively reviewing related studies on detecting and describing two-dimensional image line segments to provide researchers with an overall picture and deep understanding. Based on their mechanisms, two taxonomies for line segment detection and description are presented to introduce, analyze, and summarize these studies, facilitating researchers to learn about them quickly and extensively. The key issues, core ideas, advantages and disadvantages of existing methods, and their potential applications for each category are analyzed and summarized, including previously unknown findings. The challenges in existing methods and corresponding insights for potentially solving them are also provided to inspire researchers. In addition, some state-of-the-art line segment detection and description algorithms are evaluated without bias, and the evaluation code will be publicly available. The theoretical analysis, coupled with the experimental results, can guide researchers in selecting the best method for their intended vision applications. Finally, this study provides insights for potentially interesting future research directions to attract more attention from researchers to this field.

Figures (14)

• Figure 1: Local image features on a test image: (a) Interest points FASTER, (b) Line segments EDLines, and (c) Corresponding edge map EdgeDrawing of line segments in (b). Line segments contain more structural and geometric features about a scene than points and are more compact in data expression than edges.
• Figure 2: The structure of this review and high-level taxonomies of existing line segment detection and description methods.
• Figure 3: (a) - (b): The example in ExtendedHoughtransformforlinearfeaturedetection shows the ideal Hough transformation in the Cartesian coordinate system. Two points $A(x_n, y_n)$ and $B(x_m, y_m)$ on a line $y=\alpha_lx+\beta_l$ in the image space correspond to two lines $\beta=-\alpha x_n + y_n$ and $\beta=-\alpha x_m + y_m$ in the Hough space, and they intersect at a point $C(\alpha_l, \beta_l)$. (c) - (d): The example shows the ideal Hough transformation in the polar coordinate system. Four points on a line in the image space correspond to four curves in the Hough space, and they intersect at a point.
• Figure 4: Visual illustrations in the LSD LSDLSDaLineSegmentDetector algorithm: (a) image region of interest; (b) level lines (small black lines) and the approximated rectangles for line ROSs. Line segments are detected from these approximated rectangles.
• Figure 5: In the EDLines EDLines algorithm, line segments (red lines) are detected from edges (green points). The figure shows a test image (a) and a region of interest (b) with drawn edges and detected line segments.