An algorithm for motion prediction using a biological visual sensor

An algorithm for translational motion prediction is presented. The algorithm predicts motion without using correspondence and using only one line correlation instead of full image correlation. The algorithm was developed in conjunction with research on sensors for machine vision with biological vision features. Tests with noisy synthetic images produced good results.