2D Generalized S-Transform Program with Test Suite

This package contains both the 2D generalized S-transform implementation and corresponding test programs designed for performing time-frequency analysis on images. The implementation uses MATLAB R2011b running on 32-bit Windows 7 systems. The core algorithm implements a two-dimensional extension of the S-transform, employing frequency-adaptive Gaussian windows that provide optimal time-frequency resolution trade-offs. The computational implementation involves nested loops for frequency domain convolution operations with memory-intensive matrix operations. System requirements: 2GB RAM configuration currently limits processing to maximum 84x84 pixel images due to memory constraints in the matrix computation phase. Larger images will trigger 'out of memory' errors primarily because the algorithm requires storing complex-valued frequency matrices whose size grows quadratically with image dimensions. Key functions include: - gst_2d() - Main transformation function implementing the 2D S-transform with configurable window parameters - test_gst() - Validation script demonstrating proper usage and performance benchmarks - visualize_tf() - Time-frequency visualization utility with interactive plotting capabilities This represents the first publicly available implementation of 2D generalized S-transform, currently unavailable on major code repositories including this site, PUDN, and ilovematlab. We encourage community contributions by uploading to these three platforms and collaborating on algorithm optimization, particularly for memory management and computational efficiency improvements.