1. Software version
matlab2021a
2. Theoretical knowledge of this algorithm
Figure shows a flowchart of the gait recognition research strategy. Throughout the research, we will record walking videos of multiple people from real-world scenarios. The human target is extracted from the walking video by using the color model, and the binary image sequence of the human target walking is obtained. Then, the GEI is calculated from the binary image sequence, the GEI images are divided into training set and test set, and the GEI training set is trained by CNN. Finally, we will use the trained CNN model to test the recognition accuracy on the test set.
3. Core code
alexnet
clc;
clear;
close all;
warning off;
rng(1);
SEL = 12;
if SEL == 1
digitDatasetPath = ['步态能量图0t0\'];
imds = imageDatastore(digitDatasetPath,'IncludeSubfolders', true, 'LabelSource', 'foldernames');
%划分数据为训练集合验证集,训练集中每个类别包含1张图像,验证集包含其余图像的标签
numTrainFiles = 2;%设置每个类别的训练个数
[imdsTrain, imdsValidation]