Jiapeng Feng

Dr. Jiapeng Feng

orcid

Research Focus & Metrics

Academic Career Length

3Years (2024-Present)

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

Representative Publications

View All Publications
29 January 2024 in Remote Sensing
Verified
Journal Article
Open access
PDF file
XCXin Chen
Xin Chen
HZHao Zhang
Hao Zhang
SZShenghao Zhang
Shenghao Zhang
JFJiapeng Feng
Jiapeng Feng
HXHui Xia
Hui Xia
JAJianliang Ai
Jianliang Ai
Space infrared dim target recognition is an important applications of space situational awareness (SSA). Due to the weak observability and lack of geometric texture of the target, it may be unreliable to rely only on grayscale features for recognition. In this paper, an intelligent information decision-level fusion method for target recognition which takes full advantage of the ensemble classifier and Dempster–Shafer (DS) theory is proposed. To deal with the problem that DS produces counterintuitive results when evidence conflicts, a contraction–expansion function is introduced to modify the body of evidence to mitigate conflicts between pieces of evidence. In this method, preprocessing and feature extraction are first performed on the multi-frame dual-band infrared images to obtain the features of the target, which include long-wave radiant intensity, medium–long-wave radiant intensity, temperature, emissivity–area product, micromotion period, and velocity. Then, the radiation intensities are fed to the random convolutional kernel transform (ROCKET) architecture for recognition. For the micromotion period feature, a support vector machine (SVM) classifier is used, and the remaining categories of the features are input into the long short-term memory network (LSTM) for recognition, respectively. The posterior probabilities corresponding to each category, which are the result outputs of each classifier, are constructed using the basic probability assignment (BPA) function of the DS. Finally, the discrimination of the space target category is implemented according to improved DS fusion rules and decision rules. Continuous multi-frame infrared images of six flight scenes are used to evaluate the effectiveness of the proposed method. The experimental results indicate that the recognition accuracy of the proposed method in this paper can reach 93% under the strong noise level (signal-to-noise ratio is 5). Its performance outperforms single-feature recognition and other benchmark algorithms based on DS theory, which demonstrates that the proposed method can effectively enhance the recognition accuracy of space infrared dim targets.

12 November 2026 in Applied Sciences
Verified
Journal Article
Open access
PDF file
JFJiapeng Feng
Jiapeng Feng
SDShihan Dong
Shihan Dong
YLYuqing Li
Yuqing Li
SGSensen Guo
Sensen Guo
HHHaiying Hu
Haiying Hu
Unknown thrust degrades space-target tracking, while laser ranging consumes measurement time. We study the trade-off between ranging occupancy, average tracking error, and transient error using optical alerts, adaptive estimation, and guarded ranging. The multiscale-consistency-driven guarded prediction (MSC-GP) configuration is evaluated as one design within a hypothetical sensor interface. An 800 km observer tracks 60 s thrust events with 100 microradian angular RMSE, 50 m range RMSE, and synchronous 5–20 Hz effective measurements. In the original 10 Hz middle-tier comparison with the refined periodic baseline, savings are 19.60% and 18.63% over the 670 s post-initialization interval, and 9.66% and 9.17% over the full 700 s arc, for original and held-out targets, respectively. Both controls have no observed violations in these samples. A further 64 disjoint targets test six commonly tuned controller families at 10 Hz. On the nominal middle-tier confirmation set, MSC-GP uses 23.34 s of post-initialization ranging with 95.80 m RMSE; a six-second optimization baseline uses 24.29 s with 85.90 m RMSE. Neither violates the limits in 192 trajectories. SS8 remains competitive, and longer or impulsive maneuvers expose failures. These results support conditional error–occupancy trade-offs, not universal multiscale superiority or a probabilistic safety guarantee. Occupancy is effective time NL/f, not energy or hardware busy time.

28 October 2026 in Aerospace
Verified
Journal Article
Open access
PDF file
JFJiapeng Feng
Jiapeng Feng
YLYuqing Li
Yuqing Li
SGSensen Guo
Sensen Guo
WZWei Zhao
Wei Zhao
XWXuhai Wang
Xuhai Wang
JZJun Zhou
Jun Zhou
Tracking noncooperative space-object clusters couples observation duration, revisit timing, and shared constellation resources. We develop a scheduler that constrains position–velocity covariance throughout a 200 s measurement-free interval after each observation. Variable-duration visits combine time-dependent common-field acquisition, serial laser ranging, and independent target posteriors. A hierarchical differential-evolution search coordinates these visits. Nonlinear truth and an extended Kalman filter evaluate all targets in eight synthetic scenes with 72 observers and 200 Monte Carlo realizations per target. Six matched methods, five paired search seeds, and duration and budget scans separate scheduling effects. Under the reference condition, initialization-priority adaptive duration (IP-AD) gives position and velocity root-mean-square errors of 18,443.8 ± 1432.6 m and 57.102 ± 6.491 m/s. Uncertainty is the sample standard deviation of five eight-scene means. Relative to fixed 20 s visits with identical initialization priority and hold screening, the means decrease by 66.00% and 46.46%. However, relaxed planning thresholds combined with lower measurement noise reverse this benefit. Adaptive duration therefore offers conditional average gains; covariance-feasible plans do not ensure that every target meets the actual-error requirements.