Research Article

Validation of Data Association for Monocular SLAM

Algorithm 1

Pseudocode for HOHCT test and HOHCT hybrid recursive/iterative search.
Function ( , , , )   HOHCT-test ( , , , )
Input:
  matching observations found
   features observation prediction
   innovation covariance matrix
   observation Jacobian
Output:
  matching observations found
   features observation prediction
   innovation covariance matrix
   observation Jacobian
begin
m   Number of Matches in
hyp           // Grab all matches
if   JointCompatible (hyp, , , , ) then
  i   1
   while  i   m do   // Hypothesis reducer loop
     (hyp, )   HOHCT-Rec  (m, , , , , , , )
   if  JointCompatible (hyp, , , , ) then
    i   m
   else
    i   i + 1
    end if
   end while
  remove incompatible pairings from and
  update jacobian and matrix
end if
return ( , , , )
Function ( , )   HOHCT-Rec (m, , , , , , , )
Input:
m  size of full hypothesis
 size previously formed hypothesis
 hypothesis built through recursion
matches yet to remove
Output:
 best Hypothesis found from
best Mahalanobis distance
begin
if ( = ) or (m = ) then
  
     Mahalanobis ( , , , )
else
  
  Mahalanobis ( , , , )
  for     ( ) : (m −   + 1) do
      ( )   HOHCT-Rec (m, + 1, , , , , , )
    If (d < ) then
           d
           h
    end if
        1
        + 1
  end for
end if
return ( , )