Abstract: In order to reduce the multimedia sensor network data packet loss and time delay, improve the real-time feedback ability of monitoring network learning, proposes a multimedia sensor network optimization based on linear shift adaptive learning monitor real-time feedback method. The optimal deployment model of multimedia sensor network, the sensor nodes within the cluster fusion tracking method for multimedia information fusion and the two-dimensional entropy feature extraction, fuzzy adaptive control based on sensor node weighted threshold, combined with linear shift channel allocation method for multimedia sensor network learning monitoring feedback link equalization, using intelligent optimization algorithm for multimedia sensor network transmission delay control, improve the ability of real-time feedback data. Simulation results show that using the method of multimedia the sensor network monitoring real-time feedback control, output data lost packet rate is low, the time delay of the multimedia information feedback is small, and the link balance of the whole network is better.