Using low cost video cameras in automation: a close range photogrammetry approach

In this paper, a new calibration procedure is introduced which simplifies the estimation of systematic errors in low cost video cameras, improves their accuracy and makes them suitable for three-dimensional measurements. Photogrammetry utilizes two kinds of cameras: metric, where calibration is performed with expensive equipment (in which interior parameters and the mathematical models for lens distortion are well-defined); or nonmetric, where the systematic coordinate errors are compensated by analytical models applied directly to the photogrammetric projective equations (self calibration). These methods imply an accurate determination of dimensions of pixels while geometric degradation of the CCD array is approached by polynomial coefficients. Low cost video cameras are expected to have significant deviation of their nominal settings, making the estimation of distortions in combination with the influence of environmental parameters, a hard working problem. As is explained, a calibration method that could confront these problems in the application field, using equipment with reasonable costs, will be a good improvement