Conference PaperPDF Available

Crater monitoring through social media observations

Authors:
Ilias Gialampoukidis at The Centre for Research and Technology, Hellas
Ilias Gialampoukidis
Stefanos Vrochidis at The Centre for Research and Technology, Hellas
Stefanos Vrochidis
Ioannis (Yiannis) Kompatsiaris at The Centre for Research and Technology, Hellas
Ioannis (Yiannis) Kompatsiaris
Download full-text PDF
Read full-text
Download citation

Figures

Figures - uploaded by Ilias Gialampoukidis
Author content
All figure content in this area was uploaded by Ilias Gialampoukidis
Content may be subject to copyright.
Content uploaded by Ilias Gialampoukidis
Author content
All content in this area was uploaded by Ilias Gialampoukidis on Jun 19, 2017
Content may be subject to copyright.
Crater monitoring through social media observations
I. Gialampoukidis, S. Vrochidis and I. Kompatsiaris
Centre for Research and Technology Hellas, Information Technologies Institute (CERTH-ITI), Thessaloniki, Greece
E-mail: {heliasgj, stefanos, ikom}@iti.gr
Abstract
Lunar craters have attracted the attention of not only
scientists but also citizens. Modern high-resolution
cameras with zoom capabilities allow citizens to cap-
ture and share pictures of the Moon in Social Media
platforms, such as Twitter. We have collected 69 pic-
tures of the Moon, from 01-01-2017 to 17-04-2017,
that have been uploaded on Twitter and have been as-
sociated with the keyword #crater. The lunar pictures
are indexed using SIFT descriptors and are then clus-
tered using density-based approaches to group them
into the automatically detected levels of zoom.
1. Introduction
The automatic detection and classification of lunar
craters have been one of the most important challenges
among lunar experts. Several approaches have been
proposed [7] to detect or count craters, in order to as-
sist and accelerate the classification of space images.
Crater shapes are changing in time and their transition
to a more complex morphology has been investigated
[5]. Other approaches involve the extraction of visual
descriptors that are based on the Hough transform for
the detection and counting of craters [3]. Contrary to
the use of lunar catalogues of optical images [6], we
propose in this work the monitoring of crater activity
through social media observations.
2. Methodology and Results
We crawled from the Twitter API169 pictures of the
Moon, from 01-01-2017 to 17-04-2017, in response to
the keyword #crater. In Table 1 we present the number
of pictures per month and the daily coverage:
Daily coverage =number of pictures per month
number of days per month
(1)
1https://dev.twitter.com
From each crawled Twitter image, we extract salient
points using the Lip-vireo2tool, in order to index
all images using SIFT descriptors [4]. The Bag-
of-Visual-Words representation is followed [2] using
term frequency - inverse document frequency (tf-idf)
scores, using a visual vocabulary of 100 visual words,
obtained by k-means clustering with 30 iterations to
ensure convergence in the visual vocabulary creation.
Table 1: Uploaded pictures per month.
Month Pictures Daily Coverage
January 23 74.19%
February 19 67.86%
March 16 51.61%
April (1st-17th) 11 64.70%
After the indexing of each image, an OPTICS reach-
ability plot [1] is employed to visualize the cluster
structure, i.e. the number of clusters and the optimal
density level. The OPTICS reachability plot indicates
that the dataset has two density-connected groups of
pictures (clusters), which are extracted at the density
level = 0.05, while a lower bound for the number of
pictures per cluster is minP ts = 5. The adoption of a
density-based clustering approach allows the presence
of noise in the dataset and does not require a priori
knowledge of the number of clusters. In Figure 1 we
present a sample of the two detected groups of lunar
images and one additional group of images (noise).
3. Summary and Conclusions
We have collected more than one image per two days,
on average, in response to the keyword #crater. Each
one of the collected images has been clustered into
two main groups of images and an additional cluster
is provided (noise) with pictures that have not been
assigned to any cluster. The proposed lunar image
2http://pami.xmu.edu.cn/ wlzhao/lip-vireo.htm
EPSC Abstracts
Vol. 11, EPSC2017-25-1, 2017
European Planetary Science Congress 2017
c
Author(s) 2017
EPSC
European Planetary Science Congress
(a) (b) (c) (d) (e) (f) (g) (h)
(i) (j) (k) (l) (m)
(n) (o) (p) (q) (r)
Figure 1: Sample of the pictures per group. The salient points that have been detected are shown on each image.
The first group of pictures (a)-(h) shows a complete facet of the Moon, while the second group (i)-(m) zooms into
craters. The last group (n)-(r) belongs to the set of unassigned pictures to any of the first two groups (noise).
clustering process provides two classes of lunar pic-
tures, at different zoom levels; the first showing a clear
view of craters grouped into one cluster and the sec-
ond demonstrating a complete view of the Moon at
various phases that are correlated with the crawling
date. The clustering stage is unsupervised, so new
topics can be detected on-the-fly. We have provided
additional sources of planetary images using crowd-
sourcing information, which is associated with meta-
data such as time, text, location, links to other users
and other related posts. This content has crater infor-
mation that can be fused with other planetary data to
enhance crater monitoring. The classification of each
new Twitter picture is marked as relevant manually,
but we plan to train a binary classifier for that purpose.
Acknowledgements
This work has supported by the EC-funded project
KRISTINA (H2020-645012).
References
[1] Ankerst, M., Breunig, M. M., Kriegel, H. P., Sander,
J.: OPTICS: ordering points to identify the clustering
structure. In ACM SIGMOD record, Vol. 28, No. 2,
pp. 49–60, 1999.
[2] Feng, J., Jiao, L. C., Zhang, X., Yang, D.: Bag-of-
visual-words based on clonal selection algorithm for
SAR image classification. IEEE Geoscience and Re-
mote Sensing Letters, Vol. 8, No. 4, pp. 691–695,
2011.
[3] Galloway, M. J., Benedix, G. K., Bland, P. A., Pax-
man, J., Towner, M. C., Tan, T.: Automated crater de-
tection and counting using the hough transform. In Im-
age Processing (ICIP), 2014 IEEE International Con-
ference on, pp. 1579–1583, 2014.
[4] Lowe, D. G.: Distinctive image features from scale-
invariant keypoints.: International journal of computer
vision Vol. 60, No. 2, pp. 91–110, 2004.
[5] Mahanti, P., Robinson, M., Povilaitis, R.: Crater
shapes in transition – classification of lunar impact
craters in the 5 km to 20 km size range, European Plan-
etary Science Congress, 16–21 October, Pasadena,
CA, USA, 2016.
[6] Salamuni´
ccar, G., Lonˇ
cari´
c, S., Grumpe, A., Wöhler,
C.: Hybrid method for crater detection based on to-
pography reconstruction from optical images and the
new LU78287GT catalogue of Lunar impact craters.
Advances in Space Research, Vol. 53, Vo. 12, pp.
1783–1797, 2014
[7] Sawabe, Y., Matsunaga, T., Rokugawa, S.: Automated
detection and classification of lunar craters using mul-
tiple approaches. Advances in Space Research, Vol.
37, No. 1, pp. 21–27, 2006.
ResearchGate has not been able to resolve any citations for this publication.
View
OPTICS: Ordering Points to Identify the Clustering Structure
Conference Paper
Full-text available
  • Jun 1999
  • SIGMOD REC
Cluster analysis is a primary method for database mining. It is either used as a stand-alone tool to get insight into the distribution of a data set, e.g. to focus further analysis and data processing, or as a preprocessing step for other algorithms operating on the detected clusters. Almost all of the well-known clustering algorithms require input parameters which are hard to determine but have a significant influence on the clustering result. Furthermore, for many real-data sets there does not even exist a global parameter setting for which the result of the clustering algorithm describes the intrinsic clustering structure accurately. We introduce a new algorithm for the purpose of cluster analysis which does not produce a clustering of a data set explicitly; but instead creates an augmented ordering of the database representing its density-based clustering structure. This cluster-ordering contains information which is equivalent to the density-based clusterings corresponding to a broad range of parameter settings. It is a versatile basis for both automatic and interactive cluster analysis. We show how to automatically and efficiently extract not only 'traditional' clustering information (e.g. representative points, arbitrary shaped clusters), but also the intrinsic clustering structure. For medium sized data sets, the cluster-ordering can be represented graphically and for very large data sets, we introduce an appropriate visualization technique. Both are suitable for interactive exploration of the intrinsic clustering structure offering additional insights into the distribution and correlation of the data.
View
Distinctive Image Features from Scale-Invariant Keypoints
Article
  • Nov 2004
This paper presents a method for extracting distinctive invariant features from images that can be used to perform reliable matching between different views of an object or scene. The features are invariant to image scale and rotation, and are shown to provide robust matching across a substantial range of affine distortion, change in 3D viewpoint, addition of noise, and change in illumination. The features are highly distinctive, in the sense that a single feature can be correctly matched with high probability against a large database of features from many images. This paper also describes an approach to using these features for object recognition. The recognition proceeds by matching individual features to a database of features from known objects using a fast nearest-neighbor algorithm, followed by a Hough transform to identify clusters belonging to a single object, and finally performing verification through least-squares solution for consistent pose parameters. This approach to recognition can robustly identify objects among clutter and occlusion while achieving near real-time performance.
View
Hybrid method for crater detection based on topography reconstruction from optical images and the new LU78287GT catalogue of Lunar impact craters
Article
  • Jun 2014
  • ADV SPACE RES
Impact craters are ubiquitous and well-studied structures of high geological relevance on the surfaces of the Earth's Moon, the terrestrial planets, the asteroids and the satellites of the outer planets. Therefore, it is not surprising that crater detection algorithms (CDAs) are one of the most studied subjects of image processing and analysis in lunar and planetary science. In this paper we are proposing a Hybrid CDA: a modified DEM (digital elevation map) reconstruction method used as a step in an existing CDA based on Hough transform. The new Hybrid CDA consists of: (1) reconstruction of topography from optical images using a shape from shading approach; (2) utilization of the DEM-based CDA; (3) correction of brightness and contrast of optical images used in order to be more suitable for evaluation of detections. An additional result of this work is a new method for evaluation of topography reconstruction algorithms, using a DEM-based CDA and an earlier approach for evaluation of CDAs. The new Hybrid CDA was tested using two Chandrayaan 1 Moon Mineralogy Mapper (M3) images and two excerpts of the Lunar Reconnaissance Orbiter (LRO) Wide Angle Camera (WAC) global optical image mosaic. As a result, the number of craters inside these four regions increased considerably from 1754 (as available in the previous LU60645GT catalogue) to 19 396 craters (as available in the resulting new LU78287GT catalogue). This confirmed the practical applicability of the new Hybrid CDA, which can be used in order to considerably extend current crater catalogues.
View
Bag-of-Visual-Words Based on Clonal Selection Algorithm for SAR Image Classification
Article
  • Aug 2011
Synthetic aperture radar (SAR) image classification involves two crucial issues: suitable feature representation technique and effective pattern classification methodology. Here, we concentrate on the first issue. By exploiting a famous image feature processing strategy, Bag-of-Visual-Words (BOV) in image semantic analysis and the artificial immune systems (AIS)'s abilities of learning and adaptability to solve complicated problems, we present a novel and effective image representation method for SAR image classification. In BOV, an effective fused feature sets for local feature representation are first formulated, which are viewed as the low-level features in it. After that, clonal selection algorithm (CSA) in AIS is introduced to optimize the prediction error of k-fold cross-validation for getting more suitable visual words from the low-level features. Finally, the BOV features are represented by the learned visual words for subsequent pattern classification. Compared with the other four algorithms, the proposed algorithm obtains more satisfactory and cogent classification experimental results.
View
Automated detection and classification of lunar craters using multiple approaches
Article
  • Dec 2006
  • ADV SPACE RES
Many missions such as Clementine and SELENE (SELenological and Engineering Explorer) take lunar images for examination. A large volume of imagery data has already been archived and much more is on the way. Extracting the necessary information from the already large and ever growing volume of data is the crucial problem that needs to be overcome.Craters are studied extensively since they provide us with the relative age of the surface unit and more information on the lunar surface geology. Manually extracting craters from lunar images is a difficult task because it requires a great deal of man power as well as specific knowledge and skills of extraction. Several automated craters detection algorithms have been developed but none is yet practical or sufficiently tested to be reliable.Our previous algorithm (Sawabe, Y., Matsunaga, T., Rokugawa, S. Automatic crater detection algorithm for the lunar surface using multiple approaches. J. Remote Sens. Soc. Jpn. 25 (2), 157–168, 2005.) was improved to enhance detection of craters in lunar images and automate crater classification. This algorithm was tested using various images for wide range of applicability. Four approaches were used with the crater detecting algorithm to find (1) “shady and sunny” patters in images with low sun angle, (2) circular features in edge images, (3) curves and circles in thinned and connected edge lines, and (4) discrete or broken circular edge lines using fuzzy Hough transform. The algorithm was applied to mare and highland images of the moon captured by Clementine and Apollo under different solar angles and spatial resolution. The new algorithm was able to detect 80% more without parameter tuning. In addition, the detected craters were classified by spectral characteristics derived from Clementine UV–Vis multi-spectral images. Finally, the lunar surface GIS was formulated which has the geological and spectral attributes automatically generated by our algorithm. It could be helpful system to analyze and recognize about the geological settings.
View
OPTICS: ordering points to identify the clustering structure
  • Jan 1999
  • 49-60
  • M Ankerst
  • M M Breunig
  • H P Kriegel
  • J Sander
Ankerst, M., Breunig, M. M., Kriegel, H. P., Sander, J.: OPTICS: ordering points to identify the clustering structure. In ACM SIGMOD record, Vol. 28, No. 2, pp. 49–60, 1999.
Crater shapes in transition -classification of lunar impact craters in the 5 km to 20 km size range
  • Oct 2016
  • 16-21
  • P Mahanti
  • M Robinson
  • R Povilaitis
Mahanti, P., Robinson, M., Povilaitis, R.: Crater shapes in transition -classification of lunar impact craters in the 5 km to 20 km size range, European Planetary Science Congress, 16-21 October, Pasadena, CA, USA, 2016.