Xiaochun Lei

Guilin University of Electronic Technology

Author Statistics

Papers

Citation

H-Index

i-10 index

Research Trends

Author Order

Document Type

Co-Authors

Xiaoyue Jiang

Nankai University

Junyan Chen

Fujian Medical University

Linjun Lu

University of Cambridge

Kaishan Song

Northeast Institute of Geography and Agroecology

Zhaoting Gong

Guilin University of Electronic Technology

Ruxiong Li

Jingdezhen Ceramic Institute

Zongming Wang

Northeast Institute of Geography and Agroecology

Xiong Tianmin

Guilin University of Electronic Technology

Lihong Zeng

303 Hospital of People's Liberation Army

Jia Du

Northeast Institute of Geography and Agroecology

Cooperative Institutions

Guilin University of Electronic Technology

Chinese Academy of Sciences

Northeast Institute of Geography and Agroecology

University of Chinese Academy of Sciences

Zhejiang University

Zhejiang Sci-Tech University

Sun Yat-sen University

University of Electronic Science and Technology of China

Shandong University

Central South University

Author Statistics

Papers

Citation

H-Index

i-10 index

Research Field

Application of the SSD Algorithm in a People Flow Monitoring System

2021 17th International Conference on Computational Intelligence and Security (CIS) (2019)

Yizhou Feng Junyan Chen Xiong Tianmin Rongfeng Chen Xinyu Li

For places where people flow concentrates, such as scenic spots, bus platforms, and shopping mall entrances, it is important to regulate and disperse the people flow where. This system uses an object detection algorithm named Single Shot Multibox Detector (abbreviated as SSD) to monitor the people flow in various public places. The system includes the following two parts: the embedded application layer, and the web application layer. The lightweight embedded device based on the SSD object detection algorithm can realize realtime monitoring statistics and detect abnormal crowd condition monitoring. It is convenient for managers of public places to know the flow of people in an area in realtime to deal with emergencies promptly and dispatch staff. Meanwhile, the system can improve people's experience and reduce the potential harm caused by overcrowding.

Overcrowding

10.1109/cis.2019.00079

Cite

Citations (6)

Utilization of a New Material-Guangyechu for Pulping Industry

Hailongjiang Pulp & Paper (2005)

Xiaochun Lei

A new fast growing materal-guangyechu was introduced, including utilization in present and its prospect.

Source

Cite

Citations (1)

Safety helmet detection method based on YOLO v4

2021 17th International Conference on Computational Intelligence and Security (CIS) (2020)

Deng Benyang Xiaochun Lei Miao Ye

In the actual production process, safety accidents caused by workers not wearing safety helmets often occur. In order to reduce safety accidents caused by wearing helmets, a helmet detection method based on improved YOLO v4 is proposed. By collecting a self-made data set of on-site construction site video, using the K-means algorithm to cluster the data set in getting appropriately a priori frame dimensional center and obtaining more targeted edge information. Subsequently, a multi-scale training strategy is used in the network training process to improve the adaptability of the model from different scales of detection. The experimental results show that, in the helmet wearing detection task, the model mAP value reached 92.89%, the detection speed reached 15f/s, and its detection accuracy and detection speed were improved compared with YOLO v4, which satisfies the real-time requirements of the helmet detection task.

Adaptability

10.1109/cis52066.2020.00041

Cite

Citations (43)

Effect of xylanase pretreatment of wood chips on fiber separation in the CTMP refining process

BioResources (2008)

Xiaochun Lei Lu Lin Kecheng Li

The effect of xylanase treatment of eucalyptus wood chips on chip refining and fiber properties was investigated. The fiber separation region and fiber surface structure were observed with SEM, TEM, and AFM. The fiber length and fines were analyzed with a Bauer-McNett classifier and optical image analysis of flowing suspensions (FQA). The results showed that xylanase degraded and hydrolyzed some xylan in the fiber wall, thus loosening the fiber wall structure. Therefore, in the subsequent refining process, fiber separation occurred in the secondary wall. This resulted in fibers with less lignin and extractives on the surface, which will benefit the interfiber bonding.

Refining (metallurgy)

10.15376/biores.3.3.801-815

Cite

Citations (21)

Attention-Capsule Network for Low-Light Image Recognition

2022 International Joint Conference on Neural Networks (IJCNN) (2023)

Shiqi Shen Xiaoyue Jiang Xiaochun Lei Xu Wu Yuting He

Deep learning models have made extraordinary progress in recent years, but image recognition in low-light conditions has not been studied much. Due to its application prospects, low-light image recognition has still received focuses. Low-light image recognition is still challenging because of the crucial feature information that is hard to mine and learned by a deep learning model. Compared with enhancing the image's brightness before recognition, recognizing low-light images by end-to-end mode straightly has a more practical sense. In this paper, we propose a novel end-to-end model named Attention-Capsule Network (ACNet) for low-light image recognition tasks. The proposed model is extend from capsule structure, and its key component is the Global-Local Attention (GLA) module. The GLA module is designed to effectively and comprehensively combine important information. It mines global perception information by global block and gets local detail information from the particular local union. This strategy can significantly improve the performance of the proposed model. In addition, this paper proposes a directional learning loss, which guides the model to extract key features of images by optimizing the error between reconstructed images and normal-light images. Our experimental results demonstrate the effectiveness and robustness of our proposed attention-capsule Network model for low-light image recognition.

Robustness

Feature (linguistics)

10.1109/ijcnn54540.2023.10191264

Cite

Citations (0)

Design of People Flow Monitoring System in Public Place based on MD-MCNN

Journal of Physics Conference Series (2020)

Xiong Tianmin Xiaochun Lei Zheng Xingwen Junyan Chen Yizhou Feng

Abstract Due to the limitation of hardware resources, the traditional people flow monitoring system based on computer vision in public places can’t meet different crowd-scale scenarios. Therefore, a people flow monitoring system based on MD-MCNN algorithm is designed, which is an application system combining the improved SSD object detection algorithm MNSSD and MCNN density map regression algorithm. In the initial stage, the system uses MNSSD for accurate detection and counting. If the people flow gradually reaches a certain threshold, the system automatically uses MCNN to estimate people flow until the people flow falls below the threshold. Through the experimental verification, the system can realize the people flow statistics of low-density and high-density people in different scenarios, and can be applied on the existing embedded platform. This scheme can be extended to smart cities, smart scenic spots, smart transportation and other fields.

10.1088/1742-6596/1606/1/012012

Cite

Citations (3)

Application of the Faster R-CNN Algorithm in Scene Recognition Function Design

2021 17th International Conference on Computational Intelligence and Security (CIS) (2019)

Xinyu Li Xiaochun Lei Rongfeng Chen Yizhou Feng Xiong Tianmin

The aim of smart tourism is to use information technology to improve the travel experience. The scene recognition proposed in this paper can enhance the self-guided tour experience, allowing visitors to understand the background and culture of the scenery better, thus achieving in-depth tourism and cultural tourism. Compared with other deep learning algorithms, the Faster R-CNN algorithm can improve the accuracy of scene recognition while ensuring speed. In this paper, the feasibility of the Faster R-CNN algorithm in the scene recognition function is verified by experiments, which has high theoretical value and commercial value.

Value (mathematics)

10.1109/cis.2019.00012

Cite

Citations (4)

STDC‐MA network for semantic segmentation

IET Image Processing (2022)

Xiaochun Lei Linjun Lu Xiaoyue Jiang Zhaoting Gong Chang Lü

Abstract Semantic segmentation is applied extensively in autonomous driving and intelligent transportation with methods that highly demand spatial and semantic information. Here, an STDC‐MA network is proposed to meet these demands. First, the STDC‐Seg structure is employed in STDC‐MA to ensure a lightweight and efficient structure. Subsequently, the feature alignment module is applied to understand the offset between high‐level and low‐level features, solving the problem of pixel offset related to upsampling on the high‐level feature map. The approach implements the effective fusion between high‐level features and low‐level features. A hierarchical multiscale attention mechanism is adopted to reveal the relationship among attention regions from two different input sizes of one image. Through this relationship, regions receiving much attention are integrated into the segmentation results, thereby reducing the unfocused regions of the input image and improving the effective utilisation of multiscale information. STDC‐MA maintains the segmentation speed as the STDC‐Seg network while improving the segmentation accuracy of small objects. STDC‐MA was verified on the validation dataset of Cityscapes. The segmentation result of STDC‐MA attained 78.32% mIOU with the input of 0.5× scale, 4.92% higher than STDC‐Seg.

10.1049/ipr2.12591

Cite

Citations (1)

Hierarchical Reconstruction on Image Sequences under Occluded Condition

Xiaochun Lei Chen Shen Xiaowen Liu Ruxiong Li

Hierarchical reconstruction on image sequences is a hot issue recently. When occlusion exists, the result of reconstruction will be affected. This paper proposes an algorithm to solve the occlusion issue on image sequences. The conjugate gradient method is used to estimate the projective depth. Then the initial model is built to judge the outlier. Meanwhile the final result is built with effective value. The method is proved to be feasible according to the experiment result.

10.1109/wgec.2009.157

Cite

Citations (0)

[Research on hyperspectral remote sensing in monitoring snow contamination concentration].

PubMed (2011)

Xuguang Tang Dianwei Liu Bai Zhang Jia Du Xiaochun Lei

Contaminants in the snow can be used to reflect regional and global environmental pollution caused by human activities. However, so far, the research on space-time monitoring of snow contamination concentration for a wide range or areas difficult for human to reach is very scarce. In the present paper, based on the simulated atmospheric deposition experiments, the spectroscopy technique method was applied to analyze the effect of different contamination concentration on the snow reflectance spectra. Then an evaluation of snow contamination concentration (SCC) retrieval methods was conducted using characteristic index method (SDI), principal component analysis (PCA), BP neural network and RBF neural network method, and the estimate effects of four methods were compared. The results showed that the neural network model combined with hyperspectral remote sensing data could estimate the SCC well.

Source

Cite

Citations (3)