Intelligent Recognition Method of Fuzzy Features of Surveillance Video Image Based on Big Data

In order to solve the problem of high error rate of fuzzy feature extraction in monitoring video image based on traditional fuzzy BP fuzzy feature extraction method, a fuzzy feature extraction algorithm for monitoring video image based on big data is proposed. The matched filtering method is used to reduce the noise of the monitored video image, the adaptive association rule learning algorithm is used to construct the fuzzy feature large data set, and the edge profile feature detection algorithm is used for feature decomposition. The key action point fusion method is used to enhance the fuzzy information of the monitored video image, and the region equivalent histogram of the monitored video image is extracted. The extracted gray histogram is input into the BP neural network fuzzy feature extractor, and the adaptive association rule learning algorithm is used to optimize the monitoring video image feature clustering in the hidden layer of the fuzzy feature extractor. The fuzzy feature extraction and optimization of batch monitoring video image is realized. The simulation results show that the algorithm has good accuracy, strong ability to resist inter-class attribute disturbance, and high recall rate of monitoring video image output. The time cost of fuzzy feature extraction for monitoring video image is small.