| Class | Description |
|---|---|
| GradientPrewitt_Shared |
Prewitt implementation that shares values for horizontal and vertical gradients
|
| GradientSobel_Naive |
This implementation of the sobel edge dector is implements it in such as way that the code can be easily read
and verified for correctness, however it is much slower than it needs to be.
|
| GradientSobel_Outer |
While not as fast as
GradientSobel it a big improvement over GradientSobel_Naive and much
more readable. |
| GradientSobel_UnrolledOuter |
This is a further improvement on
GradientSobel_Outer where it reduces the number of times the array needs to be
read from by saving past reads in a local variable. |
| GradientSparsePrewitt_F32 |
Sparse computation of the Prewitt gradient operator.
|
| GradientSparsePrewitt_U8 |
Sparse computation of the Prewitt gradient operator.
|
| GradientSparseSobel_F32 |
Sparse computation of the sobel gradient operator.
|
| GradientSparseSobel_U8 |
Sparse computation of the sobel gradient operator.
|
| GradientSparseThree_F32 |
Sparse computation of the three gradient operator.
|
| GradientSparseThree_U8 |
Sparse computation of the three gradient operator.
|
| GradientSparseTwo0_F32 |
Sparse computation of the two-0 gradient operator.
|
| GradientSparseTwo0_U8 |
Sparse computation of the two-0 gradient operator.
|
| GradientSparseTwo1_F32 |
Sparse computation of the two-0 gradient operator.
|
| GradientSparseTwo1_U8 |
Sparse computation of the two-0 gradient operator.
|
| GradientThree_Share |
This is an attempt to improve the performance by minimizing the number of times arrays are accessed
and partially unrolling loops.
|
| GradientThree_Standard |
Basic implementation of
GradientThree with nothing fancy is done to improve its performance. |
| GradientTwo0_Standard |
Basic implementation of
GradientTwo0 with nothing fancy is done to improve its performance. |
| GradientTwo1_Standard |
Basic implementation of
GradientTwo1 with nothing fancy is done to improve its performance. |
| HessianSobel_Shared |
Prewitt implementation that shares values for horizontal and vertical gradients
|
| HessianThree_Standard |
Basic implementation of
HessianThree with nothing fancy is done to improve its performance. |