Multiperspective Light Field Reconstruction Method via Transfer Reinforcement Learning

Algorithm 1

Transfer reinforcement learning.

(1)	Input: multiperspective image and feature vector
(2)	The feature clustering analysis is used to quantify , and the hash value is obtained
(3)	Image feature , feature environment , and Hamming distance
(4)	Let be the similarity threshold
(5)	for frame = 1, 2, 3, do
(6)	if then
(7)	Model judgment features are not similar, so select reinforcement learning
(8)	else
	Model judgment features are similar, so select transfer learning
(9)	end