Effect of color pre-processing on color-based object detection

Color-based object detection is receiving much interest recently for the potential applications in the traffic and security fields. While Color information is precious to facilitate and accelerate the object detection process, the spectral sensitivity functions of the color camera sensor have a strong effect on the quality of image colors and their appearance in the Hue -Saturation, H-S histogram of the object surface color. In this work we study the effect of color preprocessing by using a linear spectral sharpening transform, on the quality of image colors and their detection quality. It is shown that color-based object detection is severely impaired due to the spectral overlap of camera filters. We show the performance of a new spectral sharpening method, running on real images. The image color preprocessing resulted in a significant reduction of color correlation and hence clearer image colors. The transformed colors allow for fine definition of color zones in the H-S diagram and consequently enhances the detection process.

[1]  Edison Valencia,et al.  Color image sharpening inspired by human vision models. , 2006, Applied optics.

[2]  Ying Sun,et al.  A hierarchical approach to color image segmentation using homogeneity , 2000, IEEE Trans. Image Process..

[3]  Rui J. P. de Figueiredo,et al.  A nonlinear image contrast sharpening approach based on Munsell's scale , 2006, IEEE Transactions on Image Processing.

[4]  Masahiko Yachida,et al.  Tracking people by using color information from omnidirectional images , 2002, Proceedings of the 41st SICE Annual Conference. SICE 2002..

[5]  Giovanni Ramponi,et al.  A rational unsharp masking method for TV applications , 1999, Proceedings 1999 International Conference on Image Processing (Cat. 99CH36348).

[6]  Stanley T. Birchfield,et al.  Elliptical head tracking using intensity gradients and color histograms , 1998, Proceedings. 1998 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (Cat. No.98CB36231).

[7]  Lindsay Kleeman,et al.  Fast, robust colour vision for the Monash humanoid , 2000 .

[8]  Mark S. Drew,et al.  Constrained least-squares regression in color spaces , 1997, J. Electronic Imaging.

[9]  G D Finlayson,et al.  Spectral sharpening: sensor transformations for improved color constancy. , 1994, Journal of the Optical Society of America. A, Optics, image science, and vision.

[10]  Drew,et al.  Spectral sharpening with positivity , 2000, Journal of the Optical Society of America. A, Optics, image science, and vision.

[11]  Jia-Guu Leu,et al.  Edge sharpening through ramp width reduction , 2000, Image Vis. Comput..

[12]  Chao Chen,et al.  Sensor Transforms for Invariant Image Enhancement , 2002, Color Imaging Conference.

[13]  Ja-Chen Lin,et al.  Color image sharpening by moment-preserving technique , 1995, Signal Process..

[14]  Rae-Hong Park,et al.  Color image palette construction based on the HSI color system for minimizing the reconstruction error , 1996, Proceedings of 3rd IEEE International Conference on Image Processing.

[15]  B. Kartikeyan,et al.  Band sharpening of IRS-multispectral imagery by cubic spline wavelets , 2000 .

[16]  Brian V. Funt,et al.  Color Constant Color Indexing , 1995, IEEE Trans. Pattern Anal. Mach. Intell..

[17]  M. Abdellatif,et al.  Illumination-Invariant Color Recognition for Mobile Robot Navigation , 2005 .

[18]  Graham D. Finlayson,et al.  Color in Perspective , 1996, IEEE Trans. Pattern Anal. Mach. Intell..

[19]  Brian V. Funt,et al.  Experiments in Sensor Sharpening for Color Constancy , 1998, Color Imaging Conference.

[20]  Giovanni Ramponi,et al.  Image enhancement via adaptive unsharp masking , 2000, IEEE Trans. Image Process..

[21]  G. Finlayson,et al.  COEFFICIENT CHANNELS : DERIVATION AND RELATIONSHIP TO OTHER THEORETICAL STUDIES , 1996 .