Tracking features in retinal images of adaptive optics confocal scanning laser ophthalmoscope using KLT-SIFT algorithm

With the use of adaptive optics (AO), high-resolution microscopic imaging of living human retina in the single cell level has been achieved. In an adaptive optics confocal scanning laser ophthalmoscope (AOSLO) system, with a small field size (about 1 degree, 280 μm), the motion of the eye severely affects the stabilization of the real-time video images and results in significant distortions of the retina images. In this paper, Scale-Invariant Feature Transform (SIFT) is used to abstract stable point features from the retina images. Kanade-Lucas-Tomasi(KLT) algorithm is applied to track the features. With the tracked features, the image distortion in each frame is removed by the second-order polynomial transformation, and 10 successive frames are co-added to enhance the image quality. Features of special interest in an image can also be selected manually and tracked by KLT. A point on a cone is selected manually, and the cone is tracked from frame to frame.

[1]  Jeffrey B. Mulligan,et al.  Recovery of motion parameters from distortions in scanned images , 1997 .

[2]  T. Hebert,et al.  Adaptive optics scanning laser ophthalmoscopy. , 2002, Optics express.

[3]  Steven M. Jones,et al.  Adaptive-optics optical coherence tomography for high-resolution and high-speed 3D retinal in vivo imaging. , 2005, Optics express.

[4]  Donald T. Miller,et al.  Adaptive optics parallel spectral domain optical coherence tomography for imaging the living retina. , 2005, Optics express.

[5]  Yudong Zhang,et al.  Small table-top adaptive optical systems for human retinal imaging , 2002, SPIE Optics + Photonics.

[6]  Yudong Zhang,et al.  Retina imaging in vivo with the adaptive optics confocal scanning laser ophthalmoscope , 2009, Photonics and Optoelectronics Meetings.

[7]  Carlo Tomasi,et al.  Good features to track , 1994, 1994 Proceedings of IEEE Conference on Computer Vision and Pattern Recognition.

[8]  David Williams,et al.  The arrangement of the three cone classes in the living human eye , 1999, Nature.

[9]  Austin Roorda,et al.  Retinal motion estimation in adaptive optics scanning laser ophthalmoscopy. , 2006, Optics express.

[10]  Susana Marcos,et al.  Contrast improvement of confocal retinal imaging by use of phase-correcting plates. , 2002, Optics letters.

[11]  P. Artal,et al.  Adaptive-optics ultrahigh-resolution optical coherence tomography. , 2004, Optics letters.

[12]  Austin Roorda,et al.  Retinally stabilized cone-targeted stimulus delivery. , 2007, Optics express.

[13]  G LoweDavid,et al.  Distinctive Image Features from Scale-Invariant Keypoints , 2004 .

[14]  Conor Heneghan,et al.  Registration of digital retinal images using landmark correspondence by expectation maximization , 2004, Image Vis. Comput..

[15]  D R Williams,et al.  Supernormal vision and high-resolution retinal imaging through adaptive optics. , 1997, Journal of the Optical Society of America. A, Optics, image science, and vision.