python - Scipy correlate2d not normalized

Question

I'm computing an autocorrelation using scipy.signal.correlate2d

c2=sig.correlate2d(x-x.mean(), x-x.mean())

where x is a 256x256 array of float64, but am getting values greater than 1, but shouldn't correlations be between -1 and 1? Also, where can I find details of the algorithm that is being used?

Answer 1

等待大神解答