《计算机应用》唯一官方网站 ›› 2021, Vol. 41 ›› Issue (12): 3590-3595.DOI: 10.11772/j.issn.1001-9081.2021061011

• 第十八届中国机器学习会议(CCML 2021) • 上一篇    

基于灰度域特征增强的行人重识别方法

龚云鹏, 曾智勇(), 叶锋   

  1. 福建师范大学 计算机与网络空间安全学院,福州 350117
  • 收稿日期:2021-05-12 修回日期:2021-07-06 接受日期:2021-07-07 发布日期:2021-12-28 出版日期:2021-12-10
  • 通讯作者: 曾智勇
  • 作者简介:龚云鹏(1994—),男,福建泉州人,硕士研究生,主要研究方向:计算机视觉、行人重识别
    叶锋(1978—)男,福建福州人,副教授,博士,主要研究方向:计算机视觉、模式识别。

Person re-identification method based on grayscale feature enhancement

Yunpeng GONG, Zhiyong ZENG(), Feng YE   

  1. College of Computer and Cyber Security,Fujian Normal University,Fuzhou Fujian 350117,China
  • Received:2021-05-12 Revised:2021-07-06 Accepted:2021-07-07 Online:2021-12-28 Published:2021-12-10
  • Contact: Zhiyong ZENG
  • About author:GONG Yunpeng, born in 1994, M. S. candidate. His research interests include computer vison, person re-identification.
    YE Feng, born in 1978, Ph. D, associate professor. His research interests include computer vision, pattern recognition.

摘要:

在显著的类内变化中所学特征是否具有较好的不变性会决定行人重识别(ReID)模型的性能表现的上限,环境光线、图像分辨率变化、运动模糊等因素都会引起行人图像的颜色偏差,这些问题将导致模型对数据的颜色信息过度拟合从而限制模型的性能表现。而模拟数据样本的颜色信息丢失并凸显样本的结构信息可以促进模型学习到更稳健的特征。具体来说,在模型训练时,按照所设定的概率随机选择训练数据批组,然后对所选中批组中的每一个RGB图像样本随机选取图像的一个矩形区域或者直接选取整张图像,并将所选区域的像素替换为相应灰度图像中相同的矩形区域的像素,从而生成包含不同灰度区域的训练图像。实验结果表明,所提方法与基准模型相比在平均精度均值(mAP)评价指标上最高提升了3.3个百分点,并在多个数据集上表现良好。

关键词: 行人重识别, 计算机视觉, 深度学习, 数据增强, 特征鲁棒性, 全局灰度转换, 局部灰度转换

Abstract:

Whether the learned features have better invariance in the significant intra-class changes will determine the upper limit of performance of the Person Re-identification (ReID) model. Environmental light, image resolution change, motion blur and other factors may cause color deviation of pedestrian images, and these problems will cause overfitting of the model to color information of the data, thus limiting the performance of the model. By simulating the color information loss of the data samples and highlighting the structural information of the samples, the model was helped to learn more robust features. Specifically, during model training, the training batch was randomly selected according to the set probability, and then a rectangular area of the image or the entire image was randomly selected for each RGB image sample in the selected batch, and the pixels of the selected area was replaced with the pixels of the same rectangular area in the corresponding grayscale image, thus generating a training image with different grayscale areas. Experimental results demonstrate that compared with the benchmark model, the proposed method achieves a significant performance improvement of 3.3 percentage points at most on the evaluation index mean Average Precision (mAP), and performs well on multiple datasets.

Key words: person re-identification, computer vision, deep learning, data augmentation, feature robustness, Global Grayscale Transformation (GGT), Local Grayscale Transformation (LGT)

中图分类号: