Y. Kim^1, P. Ling, K. C. Ting¹, N. Kondo

Image processing algorithms were developed to recover 3-D position information that was used to locate plant leaves for non-contact leaf temperature measurement. A new algorithm was developed to measure view depth using a monocular camera equipped with a computer-controlled zoom lens. The methodology has improved depth recovery resolution over the conventional monocular imaging technique. An algorithm was also developed to find a maximum enclosed circle on a leaf surface so the field of view of an infrared temperature sensor could be fully filled by the target without peripheral noisy vision. The center of the enclosed circle and the estimated depth defined X and Y, and Z coordinates of the sensor location, respectively.

Keywords: zoom lens, depth measurement, machine vision, plant sensing, image processing, non-contact sensing