Generally, for each GCP, there is one corresponding error equation
like Eq. (4). Therefore, in practice, Eq. (4) can be extended
according to the number of images and configuration of the
ground control points and tie points. As a result, object coordinates
of tie points, as well as the bias correction parameters for
each image, can be resolved using the least squares of the bundle
adjustment on the basis of Eq. (4). For more details on the bundle
adjustment, see for example, (Grodecki and Dial, 2003) and
(Fraser and Hanley, 2003). The RMS errors for the above-mentioned
bias-compensation models are assessed on the CKPs by
calculating the difference of their coordinates computed from
the bias-compensated models and the coordinates obtained from
GPS survey.
Generally, for each GCP, there is one corresponding error equationlike Eq. (4). Therefore, in practice, Eq. (4) can be extendedaccording to the number of images and configuration of theground control points and tie points. As a result, object coordinatesof tie points, as well as the bias correction parameters foreach image, can be resolved using the least squares of the bundleadjustment on the basis of Eq. (4). For more details on the bundleadjustment, see for example, (Grodecki and Dial, 2003) and(Fraser and Hanley, 2003). The RMS errors for the above-mentionedbias-compensation models are assessed on the CKPs bycalculating the difference of their coordinates computed fromthe bias-compensated models and the coordinates obtained fromGPS survey.
การแปล กรุณารอสักครู่..