Rank: 1291 based on 113 downloads (last 30 days) and 4 files submitted
photo

Mohammad Awrangjeb

E-mail
Company/University
Monash University

Personal Profile:
Professional Interests:

 

Watch this Author's files

 

Files Posted by Mohammad View all
Updated   File Tags Downloads
(last 30 days)
Comments Rating
12 Oct 2011 Screenshot A Robust Corner Matching Technique a novel corner matching technique using corners detected by contour-based arcss detector. Author: Mohammad Awrangjeb robust corner matchin..., arcss detector, cpda detector, fast cpda detector, matching contourbased..., corner detection 34 0
  • 4.0
4.0 | 1 rating
12 Oct 2011 Screenshot Affine Resilient Curvature Scale-Space Corner Detector Apply the affine-length parameterized curvature to the CSS corner detection technique. Author: Mohammad Awrangjeb corner detection, image analysis, curvature, curvature scalespace, multiscale corner det..., affine resilient 23 1
  • 4.5
4.5 | 2 ratings
15 Jul 2010 Screenshot A Fast Corner Detector Based on the Chord-to-Point Distance Accumulation Technique Reference: A Fast Corner Detector Based on the Chord-to-Point Distance Accumulation Technique, IEEE. Author: Mohammad Awrangjeb fastcpda, cpda, corner detection, chordtopoint distance..., contourbased corner d..., fast cpda 29 1
  • 5.0
5.0 | 1 rating
25 May 2009 Screenshot Robust Image Corner Detection based on the Chord-to-Point Distance Accumulation Technique A robust scale-space corner detector that outperforms existing single and multi-scale detectors. Author: Mohammad Awrangjeb multiscale detectors, corner detection, curvature, curvature scalespace, chord scalespace, image analysis 27 6
  • 4.0
4.0 | 5 ratings
Comments and Ratings by Mohammad View all
Updated File Comments Rating
27 Apr 2009 Robust Image Corner Detection based on the Chord-to-Point Distance Accumulation Technique A robust scale-space corner detector that outperforms existing single and multi-scale detectors. Author: Mohammad Awrangjeb

Hi Mike,
Thanks for your enquiries.

1. The absolute CPDA curvature values are calculated before normalization. In the previous code I forgot to take absolute value before normalization. Absolute function will protect some normalized values to become infinite (when maxium positive curvature value is 0) or to become greater than 1 (when maximum curvature value is negative). Please apply abs() function on line 118: C3(j,1:L) = abs(accumulate_chord_distance(xs,ys,chordLen,L,curve_mode(i,:)));

Note: updated code will be uploaded soon.

2. This is just an assumption that the CPDA curvature value at a T-junction point is 10. It is just to distinguish T-junctions from actual corner locations in any later applications. This value in fact cannot be used to compare with other corners during matching.

You may set any other value for your convenience.

19 Jan 2009 Robust Image Corner Detection based on the Chord-to-Point Distance Accumulation Technique A robust scale-space corner detector that outperforms existing single and multi-scale detectors. Author: Mohammad Awrangjeb

19 Jan 2009 Robust Image Corner Detection based on the Chord-to-Point Distance Accumulation Technique A robust scale-space corner detector that outperforms existing single and multi-scale detectors. Author: Mohammad Awrangjeb

19 Jan 2009 Robust Image Corner Detection based on the Chord-to-Point Distance Accumulation Technique A robust scale-space corner detector that outperforms existing single and multi-scale detectors. Author: Mohammad Awrangjeb

Sorry, own rating! it happended accidentally. I could not removed afterwards! I did re-rate to make the previous rating back!

19 Jan 2009 Robust Image Corner Detection based on the Chord-to-Point Distance Accumulation Technique A robust scale-space corner detector that outperforms existing single and multi-scale detectors. Author: Mohammad Awrangjeb

I fixed and updated the bug today. Mathworks will show the change soon.

Bug fixing: add/update the following two lines

left = smoothed_curve(extremum(1):-1:1,:);
right = smoothed_curve(end:-1:extremum(end),:);

before the follwoing line

ang=curve_tangent([left;right],extremum(1));

Thanks to Natalia and Piao Lin.

Comments and Ratings on Mohammad's Files View all
Updated File Comment by Comments Rating
04 Apr 2014 Affine Resilient Curvature Scale-Space Corner Detector Apply the affine-length parameterized curvature to the CSS corner detection technique. Author: Mohammad Awrangjeb Rookie, Just

Hi,I want to chose the curve if its length is bigger than 15 or 20,but when I change your source code,some errors will come up.So how to change the minimum length of curve?

28 Jun 2013 Affine Resilient Curvature Scale-Space Corner Detector Apply the affine-length parameterized curvature to the CSS corner detection technique. Author: Mohammad Awrangjeb Meng

09 Jun 2013 A Fast Corner Detector Based on the Chord-to-Point Distance Accumulation Technique Reference: A Fast Corner Detector Based on the Chord-to-Point Distance Accumulation Technique, IEEE. Author: Mohammad Awrangjeb Meng

Thanks!Nice code!

17 Oct 2011 A Robust Corner Matching Technique a novel corner matching technique using corners detected by contour-based arcss detector. Author: Mohammad Awrangjeb Chong, Wan

27 Apr 2009 Robust Image Corner Detection based on the Chord-to-Point Distance Accumulation Technique A robust scale-space corner detector that outperforms existing single and multi-scale detectors. Author: Mohammad Awrangjeb Awrangjeb, Mohammad

Hi Mike,
Thanks for your enquiries.

1. The absolute CPDA curvature values are calculated before normalization. In the previous code I forgot to take absolute value before normalization. Absolute function will protect some normalized values to become infinite (when maxium positive curvature value is 0) or to become greater than 1 (when maximum curvature value is negative). Please apply abs() function on line 118: C3(j,1:L) = abs(accumulate_chord_distance(xs,ys,chordLen,L,curve_mode(i,:)));

Note: updated code will be uploaded soon.

2. This is just an assumption that the CPDA curvature value at a T-junction point is 10. It is just to distinguish T-junctions from actual corner locations in any later applications. This value in fact cannot be used to compare with other corners during matching.

You may set any other value for your convenience.

Contact us