Tumble behavior analysis and detection method based on double-flow convolutional neural network

Abstract

The invention discloses a tumble behavior analysis and detection method based on a double-flow convolutional neural network, and the method comprises the steps: firstly extracting a pedestrian contourimage and a motion history image to form a tumble data set; secondly, on the tumble data set, determining an optimal double-flow convolutional neural network structure, a spatial flow channel input contour image and a time flow channel input motion history graph by using a neural network search model; and finally, performing tumble judgment through a fusion module. The network structure adopted by the invention is the network structure with the optimal balance between precision and delay, the calculated amount caused by a redundant structure is greatly reduced, the hardware equipment cost isreduced, the analysis and detection of the falling behavior can be realized on the intelligent terminal, and the market prospect is wide.

Description

technical field [0001] The invention belongs to the field of smart home, and in particular relates to a fall behavior analysis and detection method based on a dual-stream convolutional neural network. Background technique [0002] The traditional fall behavior recognition algorithm is to manually design and extract features, and then use threshold method or machine learning algorithm to classify. However, the artificially designed features are often not comprehensive enough to accurately describe the target. Moreover, the traditional fall behavior detection needs to wear devices, such as accelerometers, gyroscopes, etc., and the operation is relatively complicated. In addition, traditional fall detection methods are difficult to scale well with more data and have poor adaptability. The development of deep learning and computer vision technology provides new solutions for fall detection. For example, convolutional neural networks can extract target features through convolut...

AI Technical Summary

Problems solved by technology

However, the effect of single-stream neural network in fall detection is not ideal, and the artificially designed network architecture often cannot achieve the trade-off between speed and accuracy, and is suboptimal

Images