A potential bug in RANSAC threshold estimation

[AtsushiHashimoto]

Hi. Thank you for your great work.

By the way, the following line may be a tiny bug that only affect thresholds for the RANSAC algorithm.

ransac_thr = thresh / np.mean([K0[0, 0], K1[1, 1], K0[0, 0], K1[1, 1]])

I thought it should be

ransac_thr = thresh / np.mean([K0[0, 0], K0[1, 1], K1[0, 0], K1[1, 1]])

It helps me a lot if @ZJUCVG could check this point.