Assessment of camera calibration towards accuracy requirement

The art of photogrammetry has expanded where non metric cameras are used to replace metric camera as technology grows. Non metric cameras are widely used today in close range photogrammetry application as the numbers of pixels increase. The non metric camera is created for photography used. The interior orientation is unknown and requires camera calibration. The research investigated on the relationship of the number of pixels and the accuracy in camera calibration and classified the accuracy based on the user's requirement. Camera calibration for optimisation accuracy and stabilisation for the non metric camera was carried out in this research. The evaluation of the accuracy of camera calibration and stability of interior orientation parameters (IOP) is evaluated by using the root mean square and standard deviation. Twelve images are taken by using the standard method in PhotoModeler Scanner Version 6.0. The numbers of photographs process are 6, 9 and 12. Processing using the default setting and preset setting were applied and the image arrangement resembles how the images are taken are altered when insert the images into the software to be processed. Results showed that the lesser the number of photographs and the number of megapixels, the smaller the root mean square error. In addition, the standard method developed by Eos Systems Inc. in terms of setting and image procedure has a smaller root mean square error. Furthermore, compact camera is inconsistent in terms of lens distortion.