M-Labs/nalgebra
3
0
Fork 1
Code Issues Pull Requests Packages Projects Releases Wiki Activity

refactor

Browse Source
This commit is contained in:
Marc Haubenstock 2019-07-28 14:16:00 +02:00
parent a62b58b529
commit 9467494ece
1 changed files with 39 additions and 58 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

97
src/linalg/convolution.rs
View File

@ -130,13 +130,12 @@ impl<N: RealField, D1: Dim, S1: Storage<N, D1>> Vector<N, D1, S1> {
} }
} }
impl<N: RealField> DMatrix<N> { impl<N: RealField> DMatrix<N> {
/// Returns the convolution of the target vector and a kernel. /// Returns the convolution of the target vector and a kernel.
/// ///
/// # Arguments /// # Arguments
/// ///
/// * `kernel` - A Matrix with rows > 0 and cols > 0 /// * `kernel` - A Matrix with rows > 0 and cols > 0 rows == cols
/// ///
/// # Errors /// # Errors
/// Inputs must satisfy `self.shape() >= kernel.shape() > 0`. /// Inputs must satisfy `self.shape() >= kernel.shape() > 0`.
@ -152,49 +151,11 @@ impl<N: RealField> DMatrix<N> {
{ {
let mat_rows = self.nrows() as i32; let mat_rows = self.nrows() as i32;
let mat_cols = self.ncols() as i32; let mat_cols = self.ncols() as i32;
let ker_rows = kernel.data.shape().0.value() as i32;
let ker_cols = kernel.data.shape().1.value() as i32;
if ker_rows == 0 || ker_rows > mat_rows || ker_cols == 0|| ker_cols > mat_cols {
panic!(
"convolve_full expects `self.nrows() >= kernel.nrows() > 0 and self.ncols() >= kernel.ncols() > 0 `, \
rows received {} and {} respectively. \
cols received {} and {} respectively.",
mat_rows, ker_rows, mat_cols, ker_cols);
}
let kernel_size = ker_rows;
let kernel_min = kernel_size/2;
let zero = zero::<N>();
let mut conv = DMatrix::<N>::zeros(mat_cols as usize, mat_rows as usize); let mut conv = DMatrix::<N>::zeros(mat_cols as usize, mat_rows as usize);
for i in 0..mat_rows {
for j in 0..mat_cols {
for k_i in 0..kernel_size {
for k_j in 0..kernel_size {
let i_matrix = i + k_i - kernel_min;
let j_matrix = j + k_j - kernel_min;
let is_i_in_range = i_matrix >=0 && i_matrix < mat_rows; convolve(&self, &kernel,&mut conv,mat_rows,mat_cols);
let is_j_in_range = j_matrix >=0 && j_matrix < mat_cols;
let convolved_value =
match is_i_in_range && is_j_in_range {
true => {
let pixel_value = *self.index((i_matrix as usize, j_matrix as usize));
let kernel_value = *kernel.index((k_i as usize,k_j as usize));
kernel_value*pixel_value
}
//TODO: More behaviour on borders
false => zero
};
*conv.index_mut((i as usize,j as usize)) += convolved_value;
}
}
}
}
conv conv
} }
@ -209,37 +170,65 @@ impl<N: RealField, R1: Dim +DimName, C1: Dim +DimName> MatrixMN<N, R1, C1> where
/// ///
/// # Arguments /// # Arguments
/// ///
/// * `kernel` - A Matrix with rows > 0 and cols > 0 /// * `kernel` - A Matrix with rows > 0 and cols > 0 and rows == cols
/// ///
/// # Errors /// # Errors
/// Inputs must satisfy `self.shape() >= kernel.shape() > 0`. /// Inputs must satisfy `self.shape() >= kernel.shape() > 0`.
/// ///
pub fn smat_convolve_full<R2, C2, S1>( pub fn smat_convolve_full<R2, C2, S2>(
&self, &self,
kernel: Matrix<N, R2, C2, S1>, //TODO: Would be nice to have an IsOdd trait. As kernels could be of even size atm kernel: Matrix<N, R2, C2, S2>, //TODO: Would be nice to have an IsOdd trait. As kernels could be of even size atm
) -> MatrixMN<N, R1, C1> ) -> MatrixMN<N, R1, C1>
where where
R2: Dim, R2: Dim,
C2: Dim, C2: Dim,
S1: Storage<N, R2, C2> S2: Storage<N, R2, C2>
{ {
let mat_rows = self.nrows() as i32; let mat_rows = self.nrows() as i32;
let mat_cols = self.ncols() as i32; let mat_cols = self.ncols() as i32;
let mut conv = MatrixMN::<N,R1,C1>::zeros();
convolve(&self, &kernel,&mut conv,mat_rows,mat_cols);
conv
}
//TODO: rest ?
}
fn convolve<N, R1, C1, R2, C2, S2>(mat: &MatrixMN<N,R1,C1>, kernel: &Matrix<N, R2, C2, S2>, target: &mut MatrixMN<N,R1,C1>, mat_rows: i32, mat_cols: i32)
where
N: RealField,
R1: Dim,
C1: Dim,
R2: Dim,
C2: Dim,
S2: Storage<N, R2, C2>,
DefaultAllocator: Allocator<N, R1, C1>
{
let ker_rows = kernel.data.shape().0.value() as i32; let ker_rows = kernel.data.shape().0.value() as i32;
let ker_cols = kernel.data.shape().1.value() as i32; let ker_cols = kernel.data.shape().1.value() as i32;
if ker_rows == 0 || ker_rows > mat_rows || ker_cols == 0|| ker_cols > mat_cols { if ker_rows == 0 || ker_rows > mat_rows || ker_cols == 0 || ker_cols > mat_cols || ker_cols != ker_rows {
panic!( panic!(
"convolve_full expects `self.nrows() >= kernel.nrows() > 0 and self.ncols() >= kernel.ncols() > 0 `, \ "convolve_full expects `self.nrows() >= kernel.nrows() > 0 and self.ncols() >= kernel.ncols() > 0 and kernel.nrows() == kernel.ncols() `, \
rows received {} and {} respectively. \ rows received {} and {} respectively. \
cols received {} and {} respectively.", cols received {} and {} respectively.",
mat_rows, ker_rows, mat_cols, ker_cols); mat_rows, ker_rows, mat_cols, ker_cols);
} }
let kernel_size = ker_rows; let kernel_size = ker_rows;
let kernel_min = kernel_size/2; let kernel_min = kernel_size/2;
let zero = zero::<N>(); let zero = zero::<N>();
let mut conv = MatrixMN::<N,R1,C1>::zeros();
for i in 0..mat_rows { for i in 0..mat_rows {
for j in 0..mat_cols { for j in 0..mat_cols {
@ -254,7 +243,7 @@ impl<N: RealField, R1: Dim +DimName, C1: Dim +DimName> MatrixMN<N, R1, C1> where
let convolved_value = let convolved_value =
match is_i_in_range && is_j_in_range { match is_i_in_range && is_j_in_range {
true => { true => {
let pixel_value = *self.index((i_matrix as usize, j_matrix as usize)); let pixel_value = *mat.index((i_matrix as usize, j_matrix as usize));
let kernel_value = *kernel.index((k_i as usize,k_j as usize)); let kernel_value = *kernel.index((k_i as usize,k_j as usize));
kernel_value*pixel_value kernel_value*pixel_value
} }
@ -262,17 +251,9 @@ impl<N: RealField, R1: Dim +DimName, C1: Dim +DimName> MatrixMN<N, R1, C1> where
false => zero false => zero
}; };
*conv.index_mut((i as usize,j as usize)) += convolved_value; *target.index_mut((i as usize,j as usize)) += convolved_value;
} }
} }
} }
} }
conv
} }
//TODO: rest ?
}