Method for geo-referencing an imaged area

摘要

A method for geo-referencing an area by an imaging optronics system which comprises acquiring M successive images by a detector, the imaged area being distributed between these M images, with M≧1. It comprises: measuring P distances d1, d2, . . . , dP between the system and P points of the area, called range-found points, with P≧3, distributed in K of said images with 1≦K≦M; acquiring the positioning xm, ym, zm of the detector at acquisition of the M images; measuring the attitude φm, θm, ψm of the detector at acquisition of the M images; acquiring the coordinates in these K images of image points (p1, q1), (p2, q2), . . . , (pP, qP) corresponding to the P range-found points; and estimating the parameters of exposure conditions xe, ye, ze, ψe, θe, φe corresponding to the M images as a function of positionings, of attitudes, distances and coordinates of the image points, to correct errors on the parameters xm, ym, zm, ψm, θm, φm of each of the M images.

主权项

1. A method for geo-referencing an area by means of an imaging optronics system which comprises a step of acquiring M successive images of the area by means of a detector, the area for which the step of acquiring is carried out being distributed between these M images, with M≧1, the method further comprising:
measuring P distances d1, d2, . . . , dp between the system and P points of the area, called range-found points, with P≧3, these range-found points being distributed in K of said images with 1≦K≦M, acquiring the positioning xm, ym, zm of the detector at the times of acquisition of the M images, measuring the attitude (φm, θm, ψm of the detector at the times of acquisition of the M images, acquiring the coordinates in these K images of the points called image points (p1, q1), (p2, q2), . . . , (pP, qP) corresponding to the P range-found points, and estimating the parameters of exposure conditions xe, ye, ze, ψe, θe, φe corresponding to the M images as a function of positionings, of attitudes, of distances and of coordinates of the image points, in order to correct the errors on the parameters xm, ym, zm, ψm, θm, φm of each of the M images, wherein M≧3, and the M images of the area are acquired in succession, these images presenting areas of overlap two by two and further comprising extracting homologous primitives in the areas of overlap of these M images and mapping the images two by two on the basis of these homologous primitives, and wherein when P=K, the range-found points are respectively at the center of each of the images.