path: root/libmoped/libs/sba-1.6/README.txt
diff options
Diffstat (limited to 'libmoped/libs/sba-1.6/README.txt')
1 files changed, 65 insertions, 0 deletions
diff --git a/libmoped/libs/sba-1.6/README.txt b/libmoped/libs/sba-1.6/README.txt
new file mode 100644
index 0000000..6744d79
--- /dev/null
+++ b/libmoped/libs/sba-1.6/README.txt
@@ -0,0 +1,65 @@
+ **************************************************************
+ version 1.6
+ By Manolis Lourakis
+ Institute of Computer Science
+ Foundation for Research and Technology - Hellas
+ Heraklion, Crete, Greece
+ **************************************************************
+==================== GENERAL ====================
+This is sba, a copylefted C/C++ implementation of generic bundle adjustment
+based on the sparse Levenberg-Marquardt algorithm. sba can support a wide
+range of manifestations/parameterizations of the multiple view reconstruction
+problem such as arbitrary projective cameras, partially or fully intrinsically
+calibrated cameras, exterior orientation (i.e. pose) estimation from fixed 3D
+points, etc. sba can be downloaded from
+sba relies on lapack for solving the augmented normal equations arising in the
+course of the Levenberg-Marquardt algorithm. if you don't already have lapack,
+I suggest getting clapack from
+Directory demo contains eucsbademo, a working example of using sba for Euclidean
+bundle adjustment.
+More details regarding sba can be found in ICS/FORTH Technical Report No. 340
+entitled "The Design and Implementation of a Generic Sparse Bundle Adjustment
+Software Package Based on the Levenberg-Marquardt Algorithm", by M.I.A. Lourakis
+and A.A. Argyros (available from
+In case that you use sba in your published work, please include a reference to
+the above TR:
+ author={M.I.A. Lourakis and A.A. Argyros},
+ title={The Design and Implementation of a Generic Sparse Bundle Adjustment Software Package
+ Based on the Levenberg-Marquardt Algorithm}
+ institution={Institute of Computer Science - FORTH},
+ address={Heraklion, Crete, Greece},
+ number={340},
+ year={2004},
+ month={Aug.},
+ note={Available from \verb+}
+==================== FILES ====================
+sba_levmar.c: SBA expert driver routines
+sba_levmar_wrap.c: simple wrappers around the routines in sba_levmar.c
+sba_lapack.c: LAPACK-based linear system solvers (LU, QR, SVD, Cholesky, Bunch-Kaufman)
+sba_crsm.c: CRS sparse matrix manipulation routines
+sba_chkjac.c: routines for verifying the correctness of user-supplied jacobians
+sba.h: Function prototypes & related data structures
+demo/*: Euclidean BA demo; see demo/README.txt for more details
+matlab/*: sba MEX-file interface; see matlab/README.txt for more details
+utils/*: Various utilities; see utils/README.txt for more details
+==================== COMPILING ====================
+ - On a Linux/Unix system, typing "make" will build both sba and the demo program.
+ - Under Windows and if Visual C is installed & configured for command line use,
+ type "nmake /f" in a cmd window to build sba and the demo program.
+ In case of trouble, read the comments on top of
+Send your comments/bug reports to