12345qiupeng
/
gtsam
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

more documentation and cleanup: missed a file

release/4.3a0
Chris Beall 2012-11-29 18:20:39 +00:00
parent 6b67238dd3
commit 879417cb0d
1 changed files with 20 additions and 15 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

33
gtsam_unstable/slam/tests/testInvDepthFactor3.cpp
View File

@ -21,6 +21,7 @@ using namespace gtsam;
static Cal3_S2::shared_ptr K(new Cal3_S2(1500, 1200, 0, 640, 480)); static Cal3_S2::shared_ptr K(new Cal3_S2(1500, 1200, 0, 640, 480));
static SharedNoiseModel sigma(noiseModel::Unit::Create(2)); static SharedNoiseModel sigma(noiseModel::Unit::Create(2));
// camera pose at (0,0,1) looking straight along the x-axis.
Pose3 level_pose = Pose3(Rot3::ypr(-M_PI/2, 0., -M_PI/2), gtsam::Point3(0,0,1)); Pose3 level_pose = Pose3(Rot3::ypr(-M_PI/2, 0., -M_PI/2), gtsam::Point3(0,0,1));
SimpleCamera level_camera(level_pose, *K); SimpleCamera level_camera(level_pose, *K);
@ -30,13 +31,16 @@ typedef NonlinearEquality<Pose3> PoseConstraint;
/* ************************************************************************* */ /* ************************************************************************* */
TEST( InvDepthFactor, optimize) { TEST( InvDepthFactor, optimize) {
// landmark 5 meters infront of camera // landmark 5 meters infront of camera (camera center at (0,0,1))
Point3 landmark(5, 0, 1); Point3 landmark(5, 0, 1);
// get expected projection using pinhole camera
Point2 expected_uv = level_camera.project(landmark); Point2 expected_uv = level_camera.project(landmark);
InvDepthCamera3<Cal3_S2> inv_camera(level_pose, K); InvDepthCamera3<Cal3_S2> inv_camera(level_pose, K);
LieVector inv_landmark(5, 1., 0., 1., 0., 0.); LieVector inv_landmark(5, 0., 0., 1., 0., 0.);
// initialize inverse depth with "incorrect" depth of 1/4
// in reality this is 1/5, but initial depth is guessed
LieScalar inv_depth(1./4); LieScalar inv_depth(1./4);
gtsam::NonlinearFactorGraph graph; gtsam::NonlinearFactorGraph graph;
@ -52,19 +56,20 @@ TEST( InvDepthFactor, optimize) {
LevenbergMarquardtParams lmParams; LevenbergMarquardtParams lmParams;
Values result = LevenbergMarquardtOptimizer(graph, initial, lmParams).optimize(); Values result = LevenbergMarquardtOptimizer(graph, initial, lmParams).optimize();
// with a single factor the incorrect initialization of 1/4 should not move!
EXPECT(assert_equal(initial, result, 1e-9)); EXPECT(assert_equal(initial, result, 1e-9));
/// Add a second camera /// Add a second camera
// add a camera 1 meter to the right // add a camera 2 meters to the right
Pose3 right_pose = level_pose * Pose3(Rot3(), Point3(2,0,0)); Pose3 right_pose = level_pose * Pose3(Rot3(), Point3(2,0,0));
SimpleCamera right_camera(right_pose, *K); SimpleCamera right_camera(right_pose, *K);
InvDepthCamera3<Cal3_S2> right_inv_camera(right_pose, K); // projection measurement of landmark into right camera
// this measurement disagrees with the depth initialization
Point3 landmark1(6,0,1); // and will push it to 1/5
Point2 right_uv = right_camera.project(landmark1); Point2 right_uv = right_camera.project(landmark);
InverseDepthFactor::shared_ptr factor1(new InverseDepthFactor(right_uv, sigma, InverseDepthFactor::shared_ptr factor1(new InverseDepthFactor(right_uv, sigma,
Symbol('x',2), Symbol('l',1),Symbol('d',1),K)); Symbol('x',2), Symbol('l',1),Symbol('d',1),K));
@ -74,16 +79,16 @@ TEST( InvDepthFactor, optimize) {
initial.insert(Symbol('x',2), right_pose); initial.insert(Symbol('x',2), right_pose);
// TODO: need to add priors to make this work with
// Values result2 = optimize<NonlinearFactorGraph>(graph, initial,
// NonlinearOptimizationParameters(),MULTIFRONTAL, GN);
Values result2 = LevenbergMarquardtOptimizer(graph, initial, lmParams).optimize(); Values result2 = LevenbergMarquardtOptimizer(graph, initial, lmParams).optimize();
Point3 l1_result2 = InvDepthCamera3<Cal3_S2>::invDepthTo3D(
Point3 result2_lmk = InvDepthCamera3<Cal3_S2>::invDepthTo3D(
result2.at<LieVector>(Symbol('l',1)), result2.at<LieVector>(Symbol('l',1)),
result2.at<LieScalar>(Symbol('d',1))); result2.at<LieScalar>(Symbol('d',1)));
EXPECT(assert_equal(landmark, result2_lmk, 1e-9));
EXPECT(assert_equal(landmark1, l1_result2, 1e-9)); // TODO: need to add priors to make this work with
// Values result2 = optimize<NonlinearFactorGraph>(graph, initial,
// NonlinearOptimizationParameters(),MULTIFRONTAL, GN);
} }