Haoxiang Lu

Dr. Haoxiang Lu


Research Focus & Metrics

Academic Career Length

7Years (2020-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
7 April 2025 in IEEE Transactions on Geoscience and Remote Sensing
Verified
Journal Article
Closed access
ZLZhenbing Liu
Zhenbing Liu
TFTianle Fang
Tianle Fang
HLHaoxiang Lu
Haoxiang Lu
WZWeidong Zhang
Weidong Zhang
RLRushi Lan
Rushi Lan
Object detection methods using deep convolutional neural networks (CNNs) have derived major advances in normal images. However, such success is hardly achieved with adverse weather due to a lack of visibility . To tackle this problem, we propose a Multi-scale Adaptive Sampling Fusion Network, named MASFNet. In this paper, we design a Feature Adaptive Enhancement Network (FAENet) consisting of three modules to adaptively perform feature enhancement on feature maps in adverse scenarios. These modules in FAENet are integrated by the Laplace pyramid, which can perform receptive field fusion, attention perception, and affine transformation for image feature enhancement. To improve the detection performance, we propose a Multi-scale Sampling Fusion Pyramid Network (MSFNet), which is capable of fusing different scale features to improve the semantic information. Experimental results demonstrate that MASFNet achieves 73.68% and 30.95% mAP on the real scene fog dataset (RTTS) and foggy driving dataset (FDD) respectively. Additionally, on the real-world scenario low illumination dataset (ExDark), MASFNet attains a substantial mAP of 63.80%, surpassing current state-of-the-art object detectors while retaining lightweight and high-speed. The source code will be released at https://github.com/PolarisFTL/MASFNet.

15 August 2023 in Applied Sciences
Verified
Journal Article
Open access
PDF file
SWShuwei Wu
Shuwei Wu
ZLZhenbing Liu
Zhenbing Liu
HLHaoxiang Lu
Haoxiang Lu
YHYingxing Huang
Yingxing Huang
Recently, object detection, which is focused on images with normal illumination levels, has achieved great success. However, the accuracy of object detection is reduced in suboptimal environments due to the images plagued by noise and low contrast. For boosting the performance of object-detection tasks under low-illumination conditions, we propose three modules for improvement: (1) the low-level feature attention (LFA) module learns to focus on the regional feature information of the object in the low-illumination environment, highlighting important features and filtering noisy information; (2) the feature fusion neck (FFN) obtains enriched feature information by fusing the feature information of the feature map after backbone; (3) the context-spatial decoupling head (CSDH) enables the classification head to focus on contextual semantic information so that the localization head obtains richer spatial details. Extensive experiments show that our algorithm realizing end-to-end detection shows good performance in low-illumination images.

14 September 2023 in Sensors
Verified
Journal Article
Open access
PDF file
WWWenhao Wang
Wenhao Wang
ZLZhenbing Liu
Zhenbing Liu
HLHaoxiang Lu
Haoxiang Lu
RLRushi Lan
Rushi Lan
ZZZhaoyuan Zhang
Zhaoyuan Zhang
Video super-resolution aims to generate high-resolution frames from low-resolution counterparts. It can be regarded as a specialized application of image super-resolution, serving various purposes, such as video display and surveillance. This paper proposes a novel method for real-time video super-resolution. It effectively exploits spatial information by utilizing the capabilities of an image super-resolution model and leverages the temporal information inherent in videos. Specifically, the method incorporates a pre-trained image super-resolution network as its foundational framework, allowing it to leverage existing expertise for super-resolution. A fast temporal information aggregation module is presented to further aggregate temporal cues across frames. By using deformable convolution to align features of neighboring frames, this module takes advantage of inter-frame dependency. In addition, it employs a hierarchical fast spatial offset feature extraction and a channel attention-based temporal fusion. A redundancy-aware inference algorithm is developed to reduce computational redundancy by reusing intermediate features, achieving real-time inferring speed. Extensive experiments on several benchmarks demonstrate that the proposed method can reconstruct satisfactory results with strong quantitative performance and visual qualities. The real-time inferring ability makes it suitable for real-world deployment.

18 August 2020 in IEEE Access
Verified
Journal Article
Open access
HLHaoxiang Lu
Haoxiang Lu
ZLZhengbing Liu
Zhengbing Liu
XPXipeng Pan
Xipeng Pan
In order to solve the problem of low contrast and fuzzy detail in infrared image, we propose an infrared image enhancement method based on multi-scale and adaptive bi-interval histogram equalization with details. The whole image enhancement method mainly consists of four parts: details enhancement, contrast stretch, edge enhancement and reconstruction of enhancement images. Firstly, the multi-scale convolution is used to enhance the details of image; Secondly, taking maximize the variance between classes and minimize the variance as fitness function and solved the threshold of the infrared image by genetic algorithm, then dividing the infrared image into two sub-intervals according to the threshold. After that, the bi-interval histogram equalization with details is applied to enhance the global contrast, at the same time, using the mean square deviation and average gray equalization to improve the brightness of the image. Finally, the enhanced image by adaptive bi-interval histogram equalization with details and the image processed by adaptive limited Laplace operator are fused by linear weighting to reconstruct the final enhancement image. The experimental results show that the proposed method can outperform state-of-the-art ones in both qualitative and quantitative comparisons.

13 November 2024 in IEEE Transactions on Instrumentation and Measurement
Verified
Journal Article
Closed access
ZLZhenbing Liu
Zhenbing Liu
JGJunming Gong
Junming Gong
HLHaoxiang Lu
Haoxiang Lu
XPXipeng Pan
Xipeng Pan
RLRushi Lan
Rushi Lan
The visualization and analysis of images captured under low-light conditions have long been a significant challenge in various fields, including surveillance, astronomy, medical imaging, and remote sensing. This work presents a new low-light image enhancement (LLIE) method called frequency features enhancement network (FFENet). We can observe that the image contains different levels of information at high and low frequencies, so we can achieve more accurate enhancement by processing the high and low frequencies separately according to their different information characteristics. Specifically, we first obtain the high- and low-frequency information of the image through the frequency decomposition (FD) block. Then, we use a contrast stretch subnetwork (CSNet) to realize the enhancement of the content of low frequency and apply a high-frequency enhancement subnetwork (HENet) to complete the detail extraction and noise suppression of high frequency. Finally, in the fusion stage, the network obtains images with normal illumination. Extensive experimental results suggest that our FFENet superiors several state-of-the-art approaches benefiting from its stronger multiscale feature expression and extraction ability. In addition, experiments also indicate that object recognition and other high-level computer applications on low-light images can be promoted owing to our designed FFENet method.

7 July 2021 in IEEE Access
Verified
Journal Article
Open access
PGPengyue Guo
Pengyue Guo
ZLZhenbing Liu
Zhenbing Liu
HLHaoxiang Lu
Haoxiang Lu
ZWZimin Wang
Zimin Wang
Hyperspectral image (HSI) contain abundant spectral and spatial information, enabling the accurate classification of ground objects. However, many existing machine learning methods have poor performance, and some existing CNN-based methods require high computational power, which considerably limits their real-world applications. To address these issues, in this paper, we propose an alternative HSI classification method based on the stacked contractive autoencoder (SCAE) and adaptive spectral-spatial information to improve the accuracy of HSI classification. Specifically, the non-subsampled shearlet transform (NSST) with the guided filtering (NG) enhances spatial structure information. Subsequently, we present an adaptive spatial information extraction method to extract the spatial information of pixels. Furthermore, we propose an HSI classification network, called SCAE-LR, for feature extraction and classification. The SCAE is implemented to extract the adaptive spectral-spatial feature, and a logistic regression (LR) layer is employed for classification. Extensive experiments on the Indian Pines data set and the Pavia University data set demonstrate the superior performance of our method.