Terrain Contour Matching (TERCOM)
TERCOM is an algorithm that analyses the mean absolute distance by matching the distance between the map depth and the measured depth, as explained by (Carreno, 2010). With TERCOM, the measurements from a certain profile along the trajectory are processed in batch, but it may also be run recursively. The method indirectly assumes constant position offsets from the INS position each time it is started.
Point Mass Filter (PMF)
PMF is a calculation that determines the position probability density function (PDF), as said by (N. Bergman, 1999). The point mass filter estimator the priori position distributions together with the error models of the depth measurements and the map to produce a posteriori probability …show more content…
This is done because real images have anomalies at some points caused by sensor errors and image deformation caused by errors of the local trajectory and residual scaled differences. The research into which type of restoration approach should be used and the equations presented which will be used was done by (M. Sistiaga, 1998). The selected method was to use Partial Derivative Equations (P.D.E.) as it successfully restored images and replaced linear filtering methods. A scheme that has been developed to restore images is multi-scale morphological analysis proposed by (L. Alvarez, 1992), which was used by (Lucido, 1997), the diffusion is based on an equation in the …show more content…
This equation means that the image is made smoother, without affecting the morphology. However, the restoration for the image is done equivalently and so it cannot tell apart homogeneous areas from uneven terrain.
This issue can also be sorted by altering the anisotropic diffusion where the research on this was done by (P. Perona, 1990). So the images can be smooth and not affect the edges, so a second directional derivative this time in parallel and orthogonal to the gradient on the image is carried out:
‘ΙΔuΙ’ relates to the gradient norm and ‘div’ to the divergence operator. The diffusion function ‘g’ is decreasing so that smoothing can be done in the homogeneous regions (ΙΔuΙ < k, k is a threshold) and stopped in a close to an edge (ΙΔuΙ > k). However, mathematic proof has been given that a function like this does not exist whilst guaranteeing system stability.
This problem was finally resolved by (Deriche, 1995) from his work, an equation was made that varies on two different parameters ‘a and b’:
where the vector ‘ξ’ now represents the orthogonal gradient and ‘η’ represents the parallel gradient. The parameters ‘a and b’ are still dependent on the diffusion function which varies with respect to the local value of the directional
TERCOM is an algorithm that analyses the mean absolute distance by matching the distance between the map depth and the measured depth, as explained by (Carreno, 2010). With TERCOM, the measurements from a certain profile along the trajectory are processed in batch, but it may also be run recursively. The method indirectly assumes constant position offsets from the INS position each time it is started.
Point Mass Filter (PMF)
PMF is a calculation that determines the position probability density function (PDF), as said by (N. Bergman, 1999). The point mass filter estimator the priori position distributions together with the error models of the depth measurements and the map to produce a posteriori probability …show more content…
This is done because real images have anomalies at some points caused by sensor errors and image deformation caused by errors of the local trajectory and residual scaled differences. The research into which type of restoration approach should be used and the equations presented which will be used was done by (M. Sistiaga, 1998). The selected method was to use Partial Derivative Equations (P.D.E.) as it successfully restored images and replaced linear filtering methods. A scheme that has been developed to restore images is multi-scale morphological analysis proposed by (L. Alvarez, 1992), which was used by (Lucido, 1997), the diffusion is based on an equation in the …show more content…
This equation means that the image is made smoother, without affecting the morphology. However, the restoration for the image is done equivalently and so it cannot tell apart homogeneous areas from uneven terrain.
This issue can also be sorted by altering the anisotropic diffusion where the research on this was done by (P. Perona, 1990). So the images can be smooth and not affect the edges, so a second directional derivative this time in parallel and orthogonal to the gradient on the image is carried out:
‘ΙΔuΙ’ relates to the gradient norm and ‘div’ to the divergence operator. The diffusion function ‘g’ is decreasing so that smoothing can be done in the homogeneous regions (ΙΔuΙ < k, k is a threshold) and stopped in a close to an edge (ΙΔuΙ > k). However, mathematic proof has been given that a function like this does not exist whilst guaranteeing system stability.
This problem was finally resolved by (Deriche, 1995) from his work, an equation was made that varies on two different parameters ‘a and b’:
where the vector ‘ξ’ now represents the orthogonal gradient and ‘η’ represents the parallel gradient. The parameters ‘a and b’ are still dependent on the diffusion function which varies with respect to the local value of the directional