FPGA implementation of decision functions
Generalized Fourier Descriptors
Curvilinear detector
Viola Jones hardware implementation
Cancer Prostate

This project is performed in collaboration with the CMP (Center for Machine Perception from CVUT, Prague, Czech Rep.), with Jiri Matas , and the Laboratory Le2i of University of Burgundy.

Cedric Lemaitre , PhD student is granted by the french MENRT, and is working on the model definition and evaluation.



Detecting specific features has been shown to be useful in many computer vision applications such as stereoscopic vision or object recognition, image retrieval. A number of feature detectors have been proposed in the literature. As these detectors can be combined in order to improve the recognition performances, it can be useful to add new features.
We propose in this work to define a new model of such features, based on the extraction of curvilinear regions, often simply named: line extraction. There are a huge number of application fields where curvilinear structures extraction is used. In photogrammetic and remote sensing tasks, it is used to extract roads, railroads or rivers from satellite or low-resolution aerial imagery, which can be used for the capture or update of data for geographic information systems. Moreover, it is useful in medical imaging for the extraction of anatomical features or simply for the segmentation of medical images from CT or MR devices.

Example of  curvilinar regions

Steger described in his IEEE PAMI paper an explicit model for lines and line profile models. A scale space analysis is carried out for each of the models. This analysis is used to derive an algorithm in which lines and their widths can be extracted with subpixel accuracy. The algorithm uses a modification of the differential approach to detect lines and their corresponding edges.
In our work, we propose an extension of Steger’s model, which allows detecting lines using all criteria of three families described above.

Publications about curvilinear detector

"Definition of a model-based detector of curvilinear regions", Cédric LEMAITRE, Johel MITERAN, Jiri MATAS, International Conference on Computer Analysis of Images and Patterns (CAIP ), Springer, Vienna, Austria, pp. 686-693, 29 August 2007.



Curvilinear detector