Abstract:The Karhunen-Loève transform (KLT) stands as a well-established discrete transform, demonstrating optimal characteristics in data decorrelation and dimensionality reduction. Its ability to condense energy compression into a select few main components has rendered it instrumental in various applications within image compression frameworks. However, computing the KLT depends on the covariance matrix of the input data, which makes it difficult to develop fast algorithms for its implementation. Approximations for the KLT, utilizing specific rounding functions, have been introduced to reduce its computational complexity. Therefore, our paper introduces a category of low-complexity, data-independent KLT approximations, employing a range of round-off functions. The design methodology of the approximate transform is defined for any block-length $N$, but emphasis is given to transforms of $N = 8$ due to its wide use in image and video compression. The proposed transforms perform well when compared to the exact KLT and approximations considering classical performance measures. For particular scenarios, our proposed transforms demonstrated superior performance when compared to KLT approximations documented in the literature. We also developed fast algorithms for the proposed transforms, further reducing the arithmetic cost associated with their implementation. Evaluation of field programmable gate array (FPGA) hardware implementation metrics was conducted. Practical applications in image encoding showed the relevance of the proposed transforms. In fact, we showed that one of the proposed transforms outperformed the exact KLT given certain compression ratios.

Fast Reciprocal Jacket Transform With Many Parameters

Fast parametric reciprocal-orthogonal jacket transforms

Fast Binary Block Inverse Jacket Transform

Fast Block Jacket Transform Based on Pauli Matrices

A Generalized Reverse Block Jacket Transform.

Person Re-identification Based on Transform Algorithm

Invariant joint transform correlator. (1) Mapping filter

Block-Circulant Inverse Orthogonal Jacket Matrices and Its Applications to the Kronecker MIMO Channel

Jacket Transform Eigenvalue Decomposition

Reversible Joint Hilbert and Linear Canonical Transform Without Distortion

Invariant joint transform correlator using coordinate transformation

A Generalised Hadamard Transform

Joint Secret Image Sharing and Progressive Transmission Based on Integer Discrete Cosine Transform

Rounded Hartley Transform: A Quasi-involution

A Novel Phase-encoded Amplitude-modulated Joint Transform Correlator

Distortion-Invariant Joint Transform Correlator Based on A Phase-Encoding Technique

RIFT2: Speeding-up RIFT with A New Rotation-Invariance Technique

Adjustable Jacobi-Fourier Moment for Image Representation

Morphological Fringe-Adjusted Joint Transform Correlation

Fast Data-independent KLT Approximations Based on Integer Functions

Multiplex Transformed Tensor Decomposition for Multidimensional Image Recovery.