Dr. Ofer Miller

Enthusiastic scientist with over 15 years of experience in research development and implementing cutting-edge technologies such as signal processing, computer vision and AI (machine learning)

• Email

HONORS AND AWARDS

• "The Most Cited Paper Award" , Image and Vision Computing Journal, published by Elsevier, June 2008.• "Celia and Marcos Maus Annual Prize" in Computer Science for distinction in Ph.D research studies, May 2002• Excellent Ph.D students scholarship from the Council for Higher Education for high technology area, December 2000.• Tel-Aviv University Award for distinction in Master of Science studies, June 1999.• Tel-Aviv University Fellowship for Master of Science students, October 1998.

Publications (How do I Contribute to Science)

[1] O. Miller, E. Navon, A. Averbuch "Tracking of Moving Objects Based on Graph edges similarity ". IEEE International Conference on Multimedia. (ICME 2003), Baltimore, USA, July 2003.[2] O. Miller, A. Pikaz, A. Averbuch "Objects based change detection in a pair of gray images", Image and Vision Computations (IVCNZ02), Auckland, November 2002.[3] A. Averbuch, Y. Keller, O. Miller, "FFTbased image registration", IEEE International Conference on Image Processing (ICIP 2002), Rochester, New York, September 2002.[4] O. Miller, A. Averbuch , "Unsupervised Segmentation of Moving MPEG Blocks Based on Classification of Temporal Information" ,Digital Image Computing (DICTA 2003), Sydney, AU.[5] O. Miller, A. Averbuch "Automatic Adaptive Segmentation of Moving Objects Based on Spatio-Temporal Information", Digital Image Computing (DICTA 2003), Sydney, AU.[6] Y. Keller, A. Averbuch, O. Miller “Robust Phase Correlation”, In Proc. of 17th International Conference On Pattern Recognition (ICPR2004), Cambridge, UK, August 2004[7] O. Miller, E. Navon, A. Averbuch, "Color Image Segmentation Based on Adaptive Local Thresholds", Elsevier , Image and Vision Computing 23 (2005) 69-85..[8] O. Miller, A. Pikaz, A. Averbuch "Objects based change detection in a pair of gray level images", Elsevier, Pattern Recognition 38 (2005) 1976-1992.[9] O. Miller, A. Averbuch, E. Navon. "Tracking of Moving Objects in Video through invariant features in their graph representation", EURASIP, Image and Video Processing Journal. (2008) , Article 328052

Now, Lets talk Computer Vision

No motion , So how can we segment the image with zero prior info ???

• X

tracking after changes between two gray images taken in different time slice , is it possible ?

Illumination independent change detection between two gray images. it does possible !

The goal of such algorithm is to extract the objects that appear only in one of two registered images. A typical application is surveillance, where a scene is sampled at different time gap. Assumption of significant illumination difference between the two images is considered. For example, one image may be captured during daylight while the other image may be captured at night with infrared device. Now the secret for the solution here is by analyzing the connectivity along gray-levels, all the blobs that are candidates to be classified as ‘change’ are extracted from both images. Then, the candidate blobs from both images are analyzed. A Blob from one image that has no matched blob in the other image is considered as a ‘change’. The algorithm is reliable, fast, accurate, and robust even under significant changes in illumination. The worst-case time complexity of the algorithm is almost linear in the image size. Therefore, it is suitable for real-time applications.

is it possible to segment accurate with no prior information and no temporal information ???

Color Image Segmentation Based on Iterative Adaptive!!!! Local Thresholds (AI is not smart enough to understand it :) )

New approach to color image segmentation. An algorithm that integrates edges and region-based techniques while local information is considered. The local consideration enables to derive local thresholds adaptively such that any threshold is associated with a specific region. As a result, the quality of the segmentation is significantly improved. The algorithm is composed of two stages. In the first stage, the watershed algorithm is applied. Its segmentation result is represented by RAG data structure and is used as an initialization for the next stage. An iterative process that derives the thresholds is the second stage. Any iteration consists of a merging process, derivation of threshold and regression process. During the merging process attributes of homogeneity of each region are saved in order to identify when inhomogeneous regions are generated. Then a threshold, which is associated with the first merge that generates inhomogeneous region, is derived. The number of thresholds is automatically determined during the process, which is also automatically terminated

• LinkedIn
• Mobile
• Email