MCA (Morphological Component Analysis) Algorithm Implementation

In this work, I implemented the MCA (Morphological Component Analysis) algorithm designed to separate point-like and curve-like targets. The implementation involved extensive development effort with continuous parameter tuning and optimization to achieve the desired separation performance. Each functional module of the algorithm is thoroughly documented with detailed descriptions and summaries, including specific code annotations explaining key functions such as morphological transformation operations, sparsity constraints, and iterative optimization procedures. The algorithm employs wavelet transforms for curve component extraction and local dictionaries for point component separation, with an iterative thresholding approach for component reconstruction. Additional optimizations and improvements have been incorporated to enhance the algorithm's robustness and applicability across various scenarios. This challenging project has yielded significant technical achievements, and I believe the comprehensive implementation details will facilitate better understanding and wider adoption of the MCA methodology.