PL-SLAM: Real-time monocular visual SLAM with points and lines
Summary (1 min read)
Introduction
- At the beginning of the 1960s GDP per capita in Spain was less than 60 percent of the average of the countries that now comprise the European Union, but by 1975 the Spanish figure was nearly 80 percent of the average.
- 1 Within this broad context, regions in Spain exhibited large differences among themselves in GDP per capita.
I. Regional Grants from the European Union and the Central Government of Spain
- Regions in Spain receive regional grants from two sources, the Spanish central government and the European Union.
- Three of these objectives are specifically aimed at regions: i) Objective 1, to promote the development and structural adjustment of less developed regions; ii) Objective 2, to help regions affected by industrial decline; and iii) Objective 5b, to help the development of agricultural areas.
- Table 1 displays these figures on a per capita basis, while Table 2 displays the figures as a percentage of GDP.
- Four additional regions, C. Valenciana, Navarra, País Vasco, and La Rioja, received relatively small amounts under the two biggest structural funds, FEDER and FCI.
II. Economic Performance of the Regions of Spain 1964-1994
- The seventeen regions of Spain, the so called Comunidades Autónomas, are quite different in their economic and social characteristics.
- Baleares, Madrid and Cataluña were the richest regions at the end of the sample, and they were among the top four at all times.
- Table 4 presents average annual growth rates of real GDP per capita.
- Beginning in mid-1970 the growth rate was slow for more than a decade until the big expansion of the second part of the 1980s when the growth rate was around four percent before declining sharply in the most recent recession.
- As can be seen from table 8, public capital has increased significantly over the period analyzed; it was barely 20 percent of GDP in 1964 and by 1991 it was nearly 40 percent.
III. Analysis of the Effect of Regional Grants on Regional Economic Performance
- To assess the impact of the grants on regional economic development, the authors compare the economic performance of two groups of regions before and after the grant policy intervention.
- The earlier time period ends in the year before the imposition of the FCI grant in Spain, while the latter period is as far into the period of both Spanish and European Union grant intervention as the data permit.
- The authors choose roughly comparable lengths for the two time periods.
- Thus, without attributing causality, there appears to be a correlation between the imposition and receipt of the grants, and an improvement in growth rates of real GDP.
Did you find this useful? Give us your feedback
Citations
329 citations
Cites methods from "PL-SLAM: Real-time monocular visual..."
...Finally, by the time of the first submission of this paper, a work with the same name (PL-SLAM, [36]) was published extending the monocular algorithm ORB-SLAM to the case of including line segment features computed through the line segment detector (LSD) detector [37]....
[...]
167 citations
165 citations
Cites background from "PL-SLAM: Real-time monocular visual..."
...For visual-only SLAM, there are several works combining point and line features to estimate camera motion [28,29]....
[...]
...Furthermore, we found that these sequences with rapid rotation caused large changes in the viewing direction, and the lighting conditions are especially challenging for tracking point features [25,26,28]....
[...]
88 citations
Cites methods from "PL-SLAM: Real-time monocular visual..."
...Based on the ORB-SLAM system, Pumarola [63] incorporated line features for designing a monocular point and line SLAM that builds a tracking model....
[...]
...on the ORB-SLAM system, Pumarola [63] incorporated line features for designing a monocular point and line SLAM that builds a tracking model....
[...]
...Based on the ORB-SLAM system, Pumarola [63] incorporated line features for designing a monocular point and line SLAM that builds a tracking model....
[...]
68 citations
References
15,558 citations
14,282 citations
"PL-SLAM: Real-time monocular visual..." refers methods in this paper
...For each possible rotation matrix we can get t1, t3 by using the trifocal tensor equations [11] which will be linear in t1, t3....
[...]
11,283 citations
"PL-SLAM: Real-time monocular visual..." refers background in this paper
...The last years have witnessed a surge in autonomous cars and aerial vehicles able to navigate for hundreds of miles without human intervention [10], [16], [32]....
[...]
8,387 citations
4,522 citations
Related Papers (5)
Frequently Asked Questions (14)
Q2. What future works have the authors mentioned in the paper "Pl-slam: real-time monocular visual slam with points and lines" ?
In future work, the authors plan to further exploit line features and incorporate other geometric primitives like planes, which can be built from lines in a similar manner as they have built lines from point features.
Q3. How do the authors preserve the realtime characteristics of ORB-SLAM?
In order to preserve the realtime characteristics of ORB-SLAM [18], the authors have carefully chosen, used and implemented fast methods for operating with lines in all stages of the pipeline: detection, triangulation, matching, culling, relocalization and optimization.
Q4. How many lines are required to solve the trifocal tensor equations?
It is worth to point that in order to get enough independent constraints when solving for the translation components using the trifocal tensor equations, the authors need two additional line correspondences, and hence, the total number of line matches required by their algorithm is five.
Q5. What is the main idea behind the proposed system?
Motivated by the need for efficient and accurate scene representations even for poorly textured environments, intasks such as visual inspection from aerial vehicles or handheld devices (i.e., with limited computational resources), the authors here propose a novel visual-based SLAM system that can combine points and lines information in a unified framework while keeping the computational cost.
Q6. Why did the authors run all the experiments five times?
Due to the randomness of the some stages of the pipeline, e.g., initialization, position optimization or global relocalization, all experiments were run five times and the authors report the median of all executions.
Q7. How are lines detected in an input frame?
Line segments in an input frame are detected by mean of LSD [31], an O(n) line segment detector, where n is the number of pixels in the image.
Q8. What are the main building blocks of the SLAM pipeline?
The authors next describe the line parameterization and error function the authors use and how this is integrated within the main building blocks of the SLAM pipeline, namely bundle adjustment, global relocalization and feature matching.
Q9. How do the authors estimate the coefficients of the j-th line?
This is achieved by means of a two-step procedure in which first minimizes the reprojection error of the detected lines and estimates the line endpoints pd,qd.
Q10. What is the way to estimate a map?
The authors next describe their line-based solution for map initialization, which can be a good alternative in low textured scenes with lack of feature points.
Q11. What is the definition of line reprojection error?
(1)The line reprojection error Eline is then defined as the sum of point-to-line distances Epl between the projected line segment endpoints, and the detected line in the image plane (see Fig. 3-right).
Q12. What is the relocalization error of the line?
In order to make their approach appropriate to handle lines for relocalization, the authors have replaced the EPnP by the recently published EPnPL [30], which minimizes the detected line reprojection error of Eq. (4).
Q13. How can dense and direct methods be used to solve this problem?
To solve this, dense and direct methods can be applied, even though they are likely to be computationally expensive [19], [21], and require dedicated GPU-implementations to achieve real-time performance.
Q14. What is the SLAM method for evaluating the localization accuracy?
To evaluate the localization accuracy the authors compare their PLSLAM method against current state-of-the-art Visual SLAM methods, including ORB-SLAM [18], PTAM [13], LSDSLAM [7] and RGBD-SLAM [6].