Color Image Processing and Applications

Pierre Gouton
Faculty of Science and Technology, Dijon, France


Abstract

This talk is organized around two research projects that are currently being conducted at the University of Burgundy, in Dijon, France: A VIDEO-BASED REAL-TIME VEHICLE COUNTING SYSTEM USING ADAPTIVE BACKGROUND METHOD This part of the talk deals with a video-based solution for real time vehicle detection and counting system, using a surveillance camera mounted on a relatively high place to acquire the traffic video stream. The two main methods applied in this system are: the adaptive background estimation and the Gaussian shadow elimination. The former allows a robust moving detection especially in complex scenes. The latter is based on color space HSV, which is able to deal with different size and intensity shadows. After these two operations, it obtains an image with moving vehicle extracted, and then operation counting is effected by a method called virtual detector. COLORIMETRIC CHARACTERIZATION OF WINES OF BURGUNDY FOR NUANCES REPRODUCTION The second part of the talk deals with the problem of colorimetric characterization of wines. As you know, the region of Burgundy in France is major wine producer. This study aims at characterizing Burgundy wines using their colors with a limited number of hues (called nuances). The study is based upon spectral transmittance measures of a large number of wine samples. Then a quantization is performed on the obtained spectra. Afterwards, the colors are classified in an appropriate color space (CIELAB) which provides a good visual uniformity and thus offers the best discrimination between the different samples. The obtained results are encouraging in the sense that they allow an accurate classification based on both spectral transmittances and hues. Hence, the proposed method allows descriptions and reproductions wine colors.