=Paper=
{{Paper
|id=Vol-2280/paper-18
|storemode=property
|title=Performance Analysis of Different Feature Detection Techniques for Modern and Old
Buildings
|pdfUrl=https://ceur-ws.org/Vol-2280/paper-18.pdf
|volume=Vol-2280
|authors=S. Rayhan Kabir,Md. Akhtaruzzaman,Rafita Haque
|dblpUrl=https://dblp.org/rec/conf/rtacsit/KabirAH18
}}
==Performance Analysis of Different Feature Detection Techniques for Modern and Old
Buildings==
https://ceur-ws.org/Vol-2280/paper-18.pdf
Performance Analysis of Different Feature Detection Techniques for
Modern and Old Buildings
S. Rayhan Kabir Md. Akhtaruzzaman Rafita Haque
Dept. of CSE Dept. of CSE Dept. of CSE
Asian University of Bangladesh Asian University of Bangladesh Asian University of Bangladesh
Dhaka, Bangladesh Dhaka, Bangladesh Dhaka, Bangladesh
rayhanhemel@gmail.com azaman01@gmail.com rafitahaque93@gmail.com
are utilizing the present application and research.
Building Detection is one of them. In recent years,
some experiments have been revealed, where computer
Abstract vision approaches are utilized in ancient architecture
and modern architecture segments.
Building detection and feature detection are
A detection technique is used in damage and
nowadays significant research fields in the
collapsed buildings, which are based on digital surface
area of computer vision. In the human eye
models [MYLY18]. Another detection method focuses
perspectives, it is very easy for separating the
on “Light Detection & Ranging” (LiDAR) method and
old and modern buildings. In the computation
detected the building by using feature compressor
aspect, differentiation of the old and modern
[NSS+18]. A manuscript has been presented a building
buildings depends on feature detection. The
detection approach using shadow, shape, and color
different building structures contain different
features of a building [GJ18]. A feature
characteristics and features. Various methods
acknowledgment method has been utilized in an ancient
of feature detection concept are being used for
structure which depends on deep learning [ZWZZ18].
collection of the features. This research paper
Here, the analysts have proposed a technique to
presents four computational methods for
distinguish the few highlights of the old structure by
detecting the feature of several modern and
utilizing a neural network system. Another ongoing
old buildings. In this experiment, we have
strategy centers on acknowledgment and perception for
analyzed Canny Edge Detection, Hough Line
antiquated Maya symbolic representation [COG18].
Transform, Find Contours and Harris Corner
After viewing the above literature review, feature
Detector techniques for the modern and old
detection of a building seems to be a very significant
buildings. After conducting these techniques,
research area and recent trends in Computer Science.
we have analyzed the performance of feature
Furthermore, these former experiments have not
detection for the modern and old buildings. In
disclosed any combined concepts about the
this manuscript, we have also shown that, why
performance of different feature detection techniques
these four techniques are suitable for detecting
for modern and old buildings. Moreover, the structures
the features of modern and old buildings.
of the modern and old buildings are not in the same
aspects. In addition, the performance or execution of
Keywords: Building Detection, Computer Vision,
feature detection techniques are displayed in different
Image Processing, Feature Detection.
activities for modern and old buildings.
According to the above research gaps, we have
1. Introduction instituted this research, where we have shown the
Object detection is right now an imperative research diverse performances of different feature detection
territory in the field of computer vision and image techniques for modern and old buildings. To construct
processing. A few kinds of identification approaches our research, we have utilized the Canny Edge Detector
[CCWT18], Hough Line Transform [TWBW18], Find
Contours [SMNC18] and Harris Corner Detector
Proceedings of the 3rd International Conference on Recent Trends and [SIV18] techniques. After utilizing these techniques,
Applications in Computer Science and Information Technology, Tiranë, we have shown different performances for the different
Albania, 23-11-2018, published at http://ceur-ws.org modern and ancient buildings. Finally, in this paper, we
have exposed a percentage rate of these four feature
detection techniques for modern and old architectures
or buildings.
2. Feature Identification for Buildings
In the computer vision aspects, there are variant types
of ideas for feature identification [GPP15], such as
corners, points, edges etc. [TG16]. In our experiment,
we have applied some techniques for collecting the
building features of modern and old dimensions.
2.1 Canny Edge Detection Original Image of Modern Building
Edge identification covers a decent variety of scientific
process that’s objectives is at distinguishing the focuses
in a picture. In our test, we have utilized the Canny
edge detection strategy. This technique was used for
recognizing an extensive variety of edges from the
picture. Various researches agree with the Canny
technique to displaying the best results in edge
detection [MA09] [KS16]. Here, the horizontal (Gx)
and vertical (Gy) directions were sifted by finding the
gradient intensity of a picture. We organized the edge
angle [MK13] for every pixel as taken after. After
implicating this approach, the gradient was always Canny Image of Modern Building
standing to edges and also rounded to the angles for the
vertical, horizontal and diagonal directions.
(1)
(2)
After implicating these equations, the gradient was
always standing to edges and also rounded to the angles
for the vertical, horizontal and diagonal directions.
Figure 1 has illustrated the output of the Canny method
for modern and old buildings and its simulation graphs. Original Image of Old Building
Canny Image of Old Building
Figure 1: Canny edge detection for different aged
buildings.
Simulation of Canny Edge Detection
2.2 Hough Line Transform
This is a feature extraction technique. It was respected
with the lines identification in a shape on the images.
Here, the line can be illuminated by two variables
[Open17]. We have denoted the variables m and b for
Cartesian coordinate method and variables r and θ for
Polar coordinate method [AOL+92]. These two
methods are utilized in Hough Line Transform
technique for identifying the line among the buildings
(See Figure 2). In our research, a line has been denoted
as y where, Original Image of Modern Building
y = mx + b (3)
In parametric form,
r = x cos θ + y sin θ. (4)
Figure 3 has been shown as the input and output of
the picture in this technique. Hence, the applied
equation of this technique is as follows:
Hough Line Transform Image of Modern Building
(5)
Original Image of Old Building
Hough Line Transform Image of Old Building
Figure 3: Hough line transform for different aged
Figure 2: Hough line transform in image buildings.
2.3 Find Contours Technique
Contours can be stated as a curve or inclination for
joining all the points’ border and having the same
color. This method was utilized for shape analysis and
object detection in a building image. In our experiment,
we have used Image Moment [ZWSP15] approach for
finding the counters of the different aged buildings. The
spatial moment of an image is denoted as mij where i
and j are nested “for loop” order. The image moment
[Open14] computed as:
Original Image of Modern Building
(6)
Figure 4 has been illustrated in several types of
counter-detection, which are used in our experiment. In
this technique, we have used cv2.findContours( )
function for stimulating the Find Contours process,
where we have denominated the inner shape as a child
and outer shape as a parent. Figure 5 has been shown as
the feature detection of buildings by using find contours
method. Find Contours Image of Modern Building
Original Image of Old Building
Image of Find Counters Method
Find Contours Image of Old Building
Figure 5: Find Contours technique for different aged
Figure 4: Find Contours theorem. buildings.
2.4 Harris Corner Detection
Corner identification is a method used to extract the
corner features of an image. In computer vision, a
corner can also be noted as a point. Harris corner
detection technique extracts the corners from an image.
It commonly finds the intensity of an image for a
prolapse of (u, v). In this approach, there is a Gaussian
window function and gives weights to pixels down. The
mathematical structure of this technique [Nelli17] is
given below which is utilized in our experiment.
Original Image of Modern Building
(7)
Here, E is the variety between the original and moved
Gaussian window. The window's dislocation in the
direction x is u and y direction is v. Window w(x, y) is
at position (x, y). The image intensity is I. Window’s
intensity is I(x+u, y+v), the original intensity is I(x, y)
and w(x, y) is a window Gaussian function. At
OpenCV, the harries corner detector function has been
entitled as cv2.cornerHarris( ). Here, Harris technique
Harris Image of Modern Building
has been improved by using its directional
differentiations and also covered the high threshold
values (See Figure 6) [CZZD09]. In Figure 7, we have
displayed the Harries approach for modern and old
buildings.
Harris Improved Harris
Original Image of Old Building
Harris Image of Old Building
Figure 8: Conner identification of different buildings by
Figure 6: Simulation of original Harris and improved
using improved Harris technique.
Harris techniques.
2. Result and Analysis True
Detection
By using Canny Edge Detection, Hough Line
Transform, Find Contours and Harris Corner Detector
techniques in modern and old aged buildings we have False
got different performances. We have done this Detection
experiment on several old and modern buildings’
images. Figure 9 has illustrated the false and true
feature detections among the images and Table 2 has
demonstrated the accuracy percentage of false and true
feature detection rates.
True True
Detection Detection
False False
Detection Detection
True and false feature detection in Find Contour images
True
Detection
True
False Detection
Detection
False
Detection
True and false feature detection in Canny images
True
Detection
False
Detection True
Detection
False
Detection
True
Detection
False True and false feature detection in Harris image
Detection
Figure 8: True and false feature detection of modern
True and false feature detection in Hough Line
Transform images and old buildings.
Table 1: Percentage of Feature Detection Accuracy References
[MYLY18] L. Moya, F. Yamazaki. , W. Liu , and M.
Feature Detection Accuracy (%) Yamada. Detection of collapsed buildings
from lidar data due to the 2016
Kumamoto earthquake in Japan. Natural
Modern Building Old Building
Hazards and Earth System Sciences,
18:65–68, January 2018.
[NSS+18] F. H. Nahhas, H. Z. M. Shafri, M. I.
Methods
Detection Type Detection Type Sameen, B. Pradhan and S. Mansor. Deep
Suitable Suitable Learning Approach for Building
True
False
True
False
Detection Using LiDAR–Orthophoto
Fusion. Journal of Sensors, 2018: article
ID 7212307, August 2018.
Canny
Edge
98% 2% Yes 98% 2% Yes [GJ18] A. J. Ghandour and A. A. Jezzini.
Detector Autonomous Building Detection Using
Edge Properties and Image Color
Invariants. Buildings, 8(5): article 65,
Hough
Line
94% 6% Yes 30% 70% No May 2018.
Transfor
m [ZWZZ18] Z. Zou, N. Wang, P. Zhao and X. Zhao.
Feature recognition and detection for
ancient architecture based on machine
Find
92% 8% Yes 90% 10% Yes
vision. In Proceedings SPIE 10602,
Contours
Method
Smart Structures and NDE for Industry
4.0, 1060209, United States, 2018.
[COG18] G. Can, J. Odobez and D. Gatica-Perez.
Harris
Corner
90% 10% Yes 95% 5% Yes How to Tell Ancient Signs Apart?
Detector Recognizing and Visualizing Maya
Glyphs with CNNs. ACM Journal on
Computing and Cultural Heritage, 1(1):
article 1, May 2018.
3. Conclusion and Future Works [CCWT18] J. Cao, L. Chen, M. Wang and Y. Tian.
Our exploration has delineated an application based Implementing a Parallel Image Edge
analysis which has shown the performance of feature Detection Algorithm Based on the Otsu-
detection techniques in feature recognition of isolated Canny Operator on the Hadoop Platform.
aged buildings. This examination is fundamentally Computational Intelligence and
centered on the period distinguishing proof by utilizing Neuroscience, 2018: article ID 3598284,
highlight location. The examination is as of now being May 2018.
worked on in the viewpoint of Deep Learning. These [TWBW18] M. Tatsubori, A. Walcott-Bryant, R.
are key research for better nearness about component Bryant, J. Wamburu. A Probabilistic
location and period recognizable proof at continuous in Hough Transform for Opportunistic
our future activities. The ongoing situations idea can be Crowd-sensing of Moving Traffic
changed into Machine Learning based model by using Obstacles. In 2018 SIAM International
Feedforward Neural Network (FNN). Conference on Data Mining, California,
USA, pages 217–215, 2018.
[SMNC18] S. Soomro, A. Munir, K. N. Choi. Hybrid brain images. In Proceedings 14th Annual
two-stage active contour method with International Conference of the IEEE
region and edge information for intensity Engineering in Medicine and Biology
inhomogeneous image segmentation. Society, pages: 1918-1919, Paris, France,
PLoS ONE, 13(1): e0191827, January 1992.
2018. [ZWSP15] Y. Zhang, S. Wang, P. Sun and P.
[SIV18] Y. Sun, E. Ientilucci, S. Voisin. Phillips. Pathological brain detection
Improvement of the Harris corner detector based on wavelet entropy and Hu moment
using an entropy-block-based strategy. In invariants. International Journal of Image
Proceedings SPIE 10644, Algorithms and Processing, 26(s1): S1283-S1290,
Technologies for Multispectral, September 2015.
Hyperspectral, and Ultraspectral [Open14] Structural Analysis and Shape
Imagery XXIV, 1064414, Florida, United Descriptors, OpenCV 2.4.13.7
States, 2018. documentation, OpenCV, 2014.
[GPP15] P. Ghosh, A. Pandey and U. C. Pati. [Nelli17] F. Nelli. OpenCV & Python – Harris
Comparison of Different Feature Corner Detection – a method to detect
Detection Techniques for Image corners in an image. Meccanismo
Mosaicing. ACCENTS Transactions on Complesso, February 2017.
Image Processing and Computer Vision,
1(1):1–7, November 2015. [CZZD09] J Chen, L. Zou, J. Zhang and L. Dou. The
Comparison and Application of Corner
[TG16] M. Thareja and A. Goyal. Performance Detection Algorithms. Journal of
Analysis of Edges, Corners and the Multimedia, 4(6):1–7, December 2009.
genres: A Subjective Estimation. IOSR
Journal of Electronics and
Communication Engineering, 1:98–104,
Conf.15010, 2016.
[MA09] R. Maini and H. Aggarwal. Study and
Comparison of Various Image Edge
Detection Techniques. International
Journal of Image Processing, 3(1):1–12,
February 2009.
[KS16] S. Kaur and I Singh. Comparison between
Edge Detection Techniques. International
Journal of Computer Applications (0975
– 8887), 145(15):15–18, July 2016.
[MK13] A. Mordvintsev and A. K. Canny Edge
Detection. OpenCV-Python Tutorials,
2013.
[Open17] Hough Line Transform, Image Processing
(imgproc module), OpenCV Tutorials,
OpenCV, 2017.
[AOL+92] J. Alakuijala, J. Oikarinen, Y. Louhisalmi,
X. Ying and J. Koivukangas. Image
transformation from polar to Cartesian
coordinates simplifies the segmentation of