Spectral Slope	Entropy	Kurtosis	Moments
1.0076	6.7895	2.0498	7.83e+09, 4.72e+25, 3.72e+41, 2.94e+57
0.8608	6.7607	0.1208	7.66e+09, 2.47e+25, 1.32e+41, 7.30e+56
0.8404	7.5126	-1.3726	2.99e+10, 7.36e+26, 2.28e+43, 7.09e+59
0.9395	7.1387	-1.4010	2.88e+10, 8.15e+26, 2.62e+43, 8.47e+59
0.0429	7.4510	0.7445	4.23e+09, 4.55e+24, 7.10e+39, 1.11e+55
0.3357	7.8067	-1.1453	2.74e+09, 7.33e+23, 2.66e+38, 9.65e+52
0.3053	7.4908	0.4256	2.17e+09, 5.27e+23, 1.62e+38, 4.99e+52

Notes on Image Analysis Metrics

2D Fast Fourier Transform (FFT)

The 2D FFT converts an image from the spatial domain to the frequency domain. In the FFT image:

The center represents low frequencies (broad features in the original image).
The edges represent high frequencies (fine details and sharp edges).
Brightness indicates the strength of the frequency component.
Horizontal and vertical lines through the center represent vertical and horizontal structures in the original image, respectively.

Power Spectrum

The power spectrum is the squared magnitude of the FFT. It shows the distribution of power (intensity) across different frequencies in the image.

Spectral Slope

The spectral slope indicates how quickly the power decreases as frequency increases. A more negative slope suggests a smoother image with less high-frequency content.

Entropy

Entropy measures the amount of information or complexity in the image. Higher entropy indicates more complex or random pixel distributions, while lower entropy suggests more uniform or predictable patterns.

Kurtosis

Kurtosis measures the "tailedness" of the pixel intensity distribution. High kurtosis indicates a distribution with heavy tails and a sharp peak, suggesting the presence of outliers or extreme values in pixel intensities.

Moments of the Power Spectrum

The moments provide statistical information about the distribution of frequencies in the image:

1st Moment: Mean power across all frequencies.
2nd Moment: Variance of the power distribution.
3rd Moment: Skewness, indicating asymmetry in the power distribution.
4th Moment: Kurtosis of the power spectrum, measuring the peakedness of the frequency distribution.

Note: The 4th moment here is the kurtosis calculated against the FFT, which differs from the kurtosis of pixel intensities in the spatial domain.

Interpreting the Results

These metrics together provide a comprehensive view of an image's characteristics:

Urban scenes often have stronger high-frequency components due to sharp edges and regular patterns.
Natural scenes typically have smoother transitions and may show more balanced frequency distributions.
Images with similar visual features often exhibit similar patterns in these metrics.