Input: Kernel predictions Z = { Z 1 , Z 2 , ⋯ , Z n }
Output: Bounding box list L
Function prediction (Z)
1) Initialize a set of zero arrays M = { M 1 , M 2 , ⋯ , M n }
2) While i = 1 to n do
3) If Z i ( p ) > δ Then M i ( p ) = T r u e // δ is a threshold with value of 0.35
4) M ← m o r p h o l o g y E x ( M )
5) K ← s c a l e E x p a n d e d ( M )
6) C , L a b e l ← c o n n e c t e d C o m p o n e n t s ( K )
7) L ← m i n A r e a R e c t B y L a b e l ( C , L a b e l )
8) return L