检索结果-内蒙古大学图书馆

Robust and Efficient Shoe Print Image Retrieval Using Spatial Transformer Network and Deep Hashing 22

学校读者我要写书评

暂无评论

Robust and Efficient Shoe Print Image Retrieval Using Spatia...

4th International Symposium on Signal Processing Systems, SSPS 2022

作者： Liu, Wei Xu, Da School of Computer Science and Technology Xi'an University of Posts and Telecommunications Key Laboratory of Impression Evidence Examination and Identification Technology Ministry of Public Security of China China School of Computer Science and Technology Xi'an University of Posts and Telecommunications China

ISBN: (纸本)9781450396103

In recent years, great progress has been made on the topic of shoe print image retrieval. However, it still remains a big challenge to accurately retrieve well-matched shoe print images from a huge database in real time. Deep hashing method has been proved to be effective in large-scale image retrieval in many applications. Output of deep hash network can be represented as a binary bit hash code, which helps to reduce storage space and retrieval time. Existing shoe print image retrieval methods have poor performance in instance retrieval because of shortage of classification information, sufficient sample information and feature quantization information. In order to overcome the problem, we put forward an end-to-end network to learn short deep hash codes. The learned hash code preserves the classification and small sample information very well. Moreover, due to the fact that STN (Spatial Transformer Network) block is simultaneously embedded into the hash network to enhance the retrieval ability for rotated shoe print images, the problem of rotation misalignment can be solved and the retrieval accuracy is improved. Furthermore, in order to make better use of class label information, we presented a new joint loss function. This loss function helps the network map both images' classification information and similarities into hash codes and reduce quantitative loss. In addition, we used triple labels to alleviate sample imbalance problem. Experiments on database including 10,500 shoe print images show that our proposed method can improve the retrieval performance. The proposed approach can yield a mAP (mean Average Precision) of 0.83 and a recall of 0.35, which demonstrates the discriminatory power of the learned hash codes in shoe print image retrieval application. © 2022 Owner/Author.

关键词： Hash functions

Sol–Gel Derived Red Fluorescent Nanocomposites of Copper Nanoclusters Conjugated Silica for Visualizing Latent Fingerprints

学校读者我要写书评

暂无评论

SSRN 2024年

作者： You, Nan Ge, Hao-Yu Yao, Xu Wang, Jiu-Jiang Li, Guo-Rui Zhang, Ying Li, Da-Wu Fan, Hongtao College of Petrochemical Engineering Liaoning Petrochemical University Fushun113001 China College of Forensic Science Criminal Investigation Police University of China Liaoning Shenyang110035 China Key Laboratory of Impression Evidence Examination and Identification Technology Ministry of Public Security Liaoning Shenyang110035 China

Phosphors with red emission have attracted increasing interests due to their high contrast and low interferences of fluorescence background which play a vital role in visualizing latent fingerprints (LFPs) in forensic science. In this work, red-emissive fluorescent nanocomposites of copper nanoclusters conjugated silica (CuNCs/silica) are prepared simply by sol-gel method in one pot for visualizing the LFPs. The CuNCs/silica exhibits the red fluorescence with an emission wavelength of 628 nm under the excitation of 292 nm and can have sufficient affinity with the LFPs ridges. The CuNCs/silica with the good photoluminescence properties can precisely visualize and capture the clear patterns and the well-defined details of the LFPs ridges on the non-porous and semi-porous surfaces of different substrates (including copper foil, compact disc, tiles, stainless steel, aluminum foil, glass slide, Hong Kong currency, colored plastic packing, colored paper and leather) even with fluorescent background. The ridge details of the aged LFPs can also be clearly identified. These suggest the great potential of the CuNCs/silica as phosphors for visualizing the LFPs in criminal investigations. © 2024, The Authors. All rights reserved.

关键词： Silica

Sol–Gel Derived Red Fluorescent Nanocomposites of Copper Nanoclusters Conjugated Silica for Visualizing Latent Fingerprints

学校读者我要写书评

暂无评论

SSRN 2024年

作者： Ge, Hao-Yu Yao, Xu Wang, Jiu-Jiang Li, Guo-Rui Li, Da-Wu You, Nan Zhang, Ying College of Petrochemical Engineering Liaoning Petrochemical University Fushun113001 China College of Forensic Science Criminal Investigation Police University of China Liaoning Shenyang110035 China Key Laboratory of Impression Evidence Examination and Identification Technology Ministry of Public Security Liaoning Shenyang110035 China

Phosphors with red emission have attracted increasing interests due to their high contrast and low background interferences which play a role in visualizing latent fingerprints (LFPs). In this work, red-emissive fluorescent nanocomposites of copper nanoclusters conjugated silica (CuNCs/silica) are prepared by sol-gel method for visualizing LFPs. The CuNCs/silica exhibits red fluorescence with an emission of 628 nm under 292 nm and has sufficient affinity with LFPs ridges. The CuNCs/silica with the good photoluminescence can precisely visualize and capture the clear patterns and well-defined details of the LFPs ridges on non-porous and semi-porous surfaces of different substrates (including copper foil, compact disc, tiles, stainless steel, aluminum foil, glass slide, Hong Kong currency, colored plastic packing, colored paper and leather) even with fluorescent background. The ridge details of the aged LFPs can be clearly identified. These suggest the great potential of the CuNCs/silica as phosphors for visualizing LFPs in criminal investigations. © 2024, The Authors. All rights reserved.

关键词： Silica

Sol–Gel Derived Red Fluorescent Nanocomposites of Copper Nanoclusters Conjugated Silica for Visualizing Latent Fingerprints

学校读者我要写书评

暂无评论

SSRN 2024年

关键词： Silica

Eye-Visible Red-Emitting Photoluminescence of Copper Nanoclusters Capped by 2-Mercaptopyrimidine Derivatives for Promoting Visualization of Latent Fingermarks

学校读者我要写书评

暂无评论

SSRN 2025年

作者： Li, Guo-Rui Dai, Ya-Ni Li, Xin-Ming Li, Dong-Xu Li, Da-Wu You, Nan Fan, Hong-Tao College of Petrochemical Engineering Liaoning Petrochemical University Fushun113001 China College of Forensic Science Criminal Investigation Police University of China Liaoning Shenyang110035 China Key Laboratory of Impression Evidence Examination and Identification Technology Ministry of Public Security Liaoning Shenyang110035 China Liaoning Petrochemical University China

Latent fingermarks (LFMs) are served as crucial evidence in screening of identification. There are only a limited number of phosphors with red emission. In this work, the copper nanoclusters (CuNCs) are prepared by chemical reduction of Cu(II) ions using 2-mercaptopyrimidine (MPY) derivatives (such as 4-methyl-2-mercaptopyrimidine (MMPY) and 4,6-dimethyl-2-mercaptopyrimidine (DMPY)), as named MMPY-CuNCs and DMPY-CuNCs, respectively. Although the MMPY and DMPY have the similar skeleton structure in molecular with MPY, the introduction of methyl groups into the skeleton of molecular plays key role in the enhancement of photoluminescence, and leads to the change in the emission wavelengths with an obvious blueshift for the asprepared CuNCs. The DMPY- CuNCs display a higher fluorescence intensity, a larger luminescence quantum yield of 60.19 % and a longer excited state lifetime of 10.14 μs compared with that of the MMPY-CuNCs. The DMPY-CuNCs with a bright red-emissive fluorescence at 700 nm under 372 nm excitation were used for the detection of LFMs left on various objects with different residual times. The DMPY-CuNCs can effectively display and accurately identify the intricate details of LFMs ridges with the features of level 1, level 2, and level 3. Non-permeable and semi-permeable surfaces of various substrates with multicolored and complicated patterns have no notable background interferences on the detection of LFMs. There is a slight drop in the image quality of the aging LFMs compared to the fresh one after the aging time of 21 days. Thus, this work opens a new avenue for the fabrication of CuCNs as a potential red-emissive phosphor that holds promise in identification and detection of the LFMs. © 2025, The Authors. All rights reserved.

关键词： Photoluminescence

A hybrid approach using human posture and contour for gait recognition under body occlusion 7

学校读者我要写书评

暂无评论

A hybrid approach using human posture and contour for gait r...

7th International Conference on Information Science and Control Engineering, ICISCE 2020

作者： Ma, Yue Wei, Chenghao Long, Hao Key Laboratory of Impression Evidence Examination and Identification Technology Ministry of Public Security China Shenzhen University Big Data Institute College of Computer Science and Software Engineering China

ISBN: (纸本)9781728164069

This paper presents a hybrid method using a Gaitset network for gait recognition. We firstly use Alphpose model to obtain key points of human posture to improve the extraction of human contour under body occlusion. The human contour map is obtained by pose2Seg model with the key points of human posture. Both human posture and body contour are used as input of the network. By using the hybrid method, it can effectively improve the accuracy and robustness of gait recognition with the upper body information under occlusion. In scenes of human walking under body occlusion, our experimental results show that the accuracy of the proposed method in self-built database can reach 84.6% accuracy. © 2020 IEEE.

关键词： Gait analysis

Nanocomposites of Nitrogen/Zinc-Doped Carbon Dots@ Hydrotalcite with Highly Fluorescent in Solid-State for Visualization of Latent Fingerprints

学校读者我要写书评

暂无评论

SSRN 2022年

作者： Li, Dawu Zhang, Xing-Fu Ding, Bao-Hong Lv, Xiao-Sen College of Forensic Science National Police University of China Liaoning Shenyang110035 China Key Laboratory of Impression Evidence Examination and Identification Technology Ministry of Public Security Liaoning Shenyang110035 China College of Petrochemical Engineering Liaoning Petrochemical University Fushun113001 China

Herein, we report on the optical performances of a newly developed nanocomposite based on N-doped carbon dots with Zn-assisted modification (Zn/N-CDs) using hydrotalcite as the support (Zn/N-CDs@hydrotalcite) for the enhanced solid-state fluorescence in its application of the visualization of latent fingerprints (LFPs). The as-prepared Zn/N-CDs@hydrotalcite can interact with universal components of the LFPs residues through electrostatic interactions and physical adsorption. Good dispersion of the Zn/N-CDs on the surface of hydrotalcite overcomes aggregation-induced quenching effect, and emits the bright green fluorescence of 525 nm under the excitation of 365 nm ultraviolet light in solid states, providing a well-defined platform for the visualization of LFPs. The feasibility of the as-prepared Zn/N-CDs@hydrotalcite as a fluorescent label for the visualization of LFPs is tested. The results illustrate that the well-defined details of ridge patterns in level 2 and level 3 of the LFPs characteristics on various substrates (such as glass, aluminum foil, tiles, plastic packing of beer bottle, paper and leather) are clearly identified for individual identification using conventional powder dusting method. Green emission of the as-prepared Zn/N-CDs@hydrotalcite can provide sufficient contrast to background interferences of substrate surfaces. The as-prepared Zn/N-CDs@hydrotalcite with high sensitivity and wide applicability has great potential as a promising alternative for the visualization of LFPs. © 2022, The Authors. All rights reserved.

关键词： Nanocomposites

New progress in the research on the stability of 3D footprint acquisition system and the application of personal identification 2020

学校读者我要写书评

暂无评论

New progress in the research on the stability of 3D footprin...

2nd International Conference on Artificial Intelligence and Advanced Manufacture, AIAM 2020

作者： Gao, Yi Key Laboratory of Evidence Examination and Identification Technology Ministry of Public Security Criminal Investigation Police University of China China

ISBN: (纸本)9781450372022

The extraction of stereoscopic field footprint has always been a major problem in the field of physical evidence processing, and the application of optical measurement technology and computer technology to extract the footprint of the site has become the main trend of footprint extraction analysis. Although the existing on-site stereoscopic footprint instruments are relatively numerous, the reliability and the stability need to be tested. Therefore, the system stability test of the 3D stereoscopic footprint acquisition instrument is a prerequisite method for instrument application. By verifying the stability of the instrument's effects, obtaining a reliable method for collecting footprints with a 3D footprint acquisition instrument. And then by experimental and statistical analysis of the stability of the three-dimensional footprint left during walking with pseudo-color pictures and indentation, summarizing the phenomenon of stable features in different footprints produced by the same person and realizing the comparison and differences of the footprints of different individuals. It seems to use the three-dimensional footprint acquisition system to collect on-site footprints or even during the subsequent detection and identification applications, which is of great significance to carry out the personal identification technology based on the biometric characteristics of the three-dimensional footprint in the future. © 2020 ACM.

关键词： System stability

Cross-Linked Curcumin/Kaolin Composite: Comprehensive Forensic Application and a Mathematical Model-Validated Method for Quantifying Fingerprint Fluorescence Contrast and Tunable Image Enhancement

学校读者我要写书评

暂无评论

SSRN 2024年

作者： Wang, Jiujiang Li, Dawu Zhang, Jianghua Gao, Zijian Han, Jinke Key Laboratory of Impression Evidence Examination and Identification Technology Ministry of Public Security College of Forensic Science Criminal Investigation Police University of China Liaoning Shenyang110854 China Key Laboratory of Anti-Drug Technology of Liaoning Provincial College of Anti-Drug and Public Security Criminal Investigation Police University of China Liaoning Shenyang110854 China Department of Information Systems Business Statistics and Operations Management HKUST Business School The Hong Kong University of Science and Technology 999077 Hong Kong Questrom School of Business Boston University BostonMA02215 United States

Fluorescent materials have unlocked promising avenues for developing latent fingerprints (LFPs), owing to their distinctive photoluminescent properties. However, existing fluorescent dyes face challenges in complex synthesis, potential toxicity, and fluorescence quenching. Meanwhile, as photoluminescent materials are increasingly used in LFP visualization, a significant gap remains in scientific methodologies to quantify fingerprint fluorescence contrast. In this study, green-emissive micron-sized curcumin/kaolin composites were synthesized via a facile one-step physical cross-linking method. Based on physical characterization, the formation of hydrogen bond network between curcumin and kaolin was verified, and the varied photoluminescent properties of the composites were clarified from the perspective of energy band. Furthermore, the curcumin/kaolin powders exhibit exceptional performance in imprint development, including high selectivity, sensitivity, versatility, stability, and contrast, realizing the development of not only LFPs on various substrates, but also knuckle prints, palm prints, and footprints, with clear three-level details. More notably, the powders successfully developed LFPs aged 30 days and even up to 100 days. Afterward, a novel method designated as fingerprint fluorescence intensity ratio (FFIR) was developed to quantify the contrast between fingerprint signal and background noise and compare the effects of different developing agents, which even achieved tunable multi-color image enhancement. The Shapiro-Wilk Normality Test-Based Paired Sample Wilcoxon Signed-Rank Test model was conducted to validate the feasibility of FFIR and its accuracy in tracking image capture parameters. The newly synthesized composites for imprint development and the mathematical model-validated method for quantifying the developed contrast and image enhancement have profound practical significance in comprehensive forensic application. © 2024, The Authors. All rights r

关键词： Hydrogen bonds