An optical-cavity-based multiwavelength sensor is developed for object discrimination and position finding. The working principle of this device employs the multiple laser beam triangulation method to determine object position in addition to its ability to recognize them. The multiwavelength sensor employs five different identically polarized and overlapped laser light beams that are sequentially pulsed and launched through a custom-made curved optical cavity to generate multiple laser spots for each laser. The intensities of the reflected light beams from the different spots are detected by a high-speed area scan image sensor. The discrimination between five different objects, namely, brick, cement sheet, roof tile, cotton, and leather is accomplished by calculating the slopes of the objects' reflectance spectra at the employed wavelengths. The object position (coordinates) are determined using the triangulation method, which is based on the projection of laser spots along determined angles on the objects and the measurements of the objects' reflectance spectra using an image sensor. Experimental results demonstrate the ability of the multiwavelength spectral reflectance sensor to simultaneously discriminate between different objects and predict their positions over a 6 m range with an accuracy exceeding 92%.
© 2011 IEEEPDF Article