No-Reference Video Quality Assessment using novel hybrid features and two-stage hybrid regression for score level fusion

This paper presents a novel No-Reference Video Quality Assessment (NR-VQA) model that utilizes proposed 3D steerable wavelet transform-based Natural Video Statistics (NVS) features as well as human perceptual features. Additionally, we proposed a novel two-stage regression scheme that significantly improves the overall performance of quality estimation. In the first stage, transform-based NVS and human perceptual features are separately passed through the proposed hybrid regression scheme: Support Vector Regression (SVR) followed by Polynomial curve fitting. The two visual quality scores predicted from the first stage are then used as features for the similar second stage. This predicts the final quality scores of distorted videos by achieving score level fusion. Extensive experiments were conducted using five authentic and four synthetic distortion databases. Experimental results demonstrate that the proposed method outperforms other published state-of-the-art benchmark methods on synthetic distortion databases and is among the top performers on authentic distortion databases. The source code is available at https://github.com/anishVNIT/two-stage-vqa.     

Simultaneous improvement of thermal conductivity and mechanical properties for mechanically mixed ABS/h-BN composites by using small amounts of hyperbranched polymer additives

Recently, thermal interface materials (TIMs) are in great demands for modern electronics. For mechanically mixed polymer composite TIMs, the thermal conductivity and the mechanical properties are generally lower than expected values due to the sharply increased viscosity and poor filler dispersion. This work shows that addition of a small amount of polyester-based hyperbranched polymer (HBP) avoided the trade-off in mechanically mixed ABS/hexagonal boron nitride (h-BN) composites. After adding 0.5 wt% HBP, the maximum h-BN content in the composites increased from 50 to 60 wt%. The out-of-plane, in-plane thermal conductivity, and tensile strength of ABS/h-BN with 50 wt% h-BN were 0.408, 0.517 W/mK, and 18 MPa, respectively, and were increased to 0.729, 0.847 W/mK, and 32 MPa by adding 0.5 wt% HBP, while 0.972, 1.12 W/mK, and 29.5 MPa were achieved for ABS/h-BN/HBP with 60 wt% h-BN. The morphological and rheological results proved that these enhancements are due to the improved h-BN dispersion by decreasing viscosity of composites during mixing. Theoretical modeling based on the modified effective medium theory confirmed such results and showed that the interfacial thermal resistance also decreased slightly. Thus, this work demonstrates a facile and scalable method for simultaneously improving the thermal conductivity and mechanical properties of thermoplastic-based TIMs.     

A signal adaptive diffusion filter for video coding: Mathematical framework and complexity reductions

In this paper we combine video compression and modern image processing methods. We construct novel iterative filter methods for prediction signals based on Partial Differential Equation (PDE) based methods. The mathematical framework of the employed diffusion filter class is given and some desirable properties are stated. In particular, two types of diffusion filters are constructed: a uniform diffusion filter using a fixed filter mask and a signal adaptive diffusion filter that incorporates the structures of the underlying prediction signal. The latter has the advantage of not attenuating existing edges while the uniform filter is less complex. The filters are embedded into a software based on HEVC with additional QTBT (Quadtree plus Binary Tree) and MTT (Multi-Type-Tree) block structure. In this setting, several measures to reduce the coding complexity of the tool are introduced, discussed and tested thoroughly. The coding complexity is reduced by up to 70% while maintaining over 80% of the gain. Overall, the diffusion filter method achieves average bitrate savings of 2.27% for Random Access having an average encoder runtime complexity of 119% and 117% decoder runtime complexity. For individual test sequences, results of 7.36% for Random Access are accomplished.     

Systematic packaging design tools integrating functional and environmental consequences on product life cycle: Case studies on laundry detergent and milk

This study presents systematic packaging design tools integrating functional and environmental consequences on product life cycle. To design packaging for sustainability, the trade-offs between functional and environmental aspects of packaging throughout the product life cycle should be considered. However, it is difficult for packaging designers to understand the overall trade-offs because the extent of the design consequences on the entire life cycle of packaging and its contents is unclear. We developed two tools for packaging design: the Life Cycle Association Matrix (LCAM) and the Function Network Diagram (FND). The following three steps, based on literature reviews and interviews with industrial experts, were applied. Firstly, we listed the product functions and design variables related to the functions as the attributes allocated to the product life cycle. Secondly, the attributes were connected appropriately based on causal relationships. Lastly, we identified the factors to support decision making in the packaging design procedure. As a result, the LCAM depicts the design consequences on the life cycle, and the FND determines the stakeholders affected by the design consequences. Two case studies were demonstrated to analyze the trade-offs by using our tools. In the case studies, a liquid laundry detergent bottle and a milk carton were redesigned. The tools identified the design consequences and stakeholders affected by the redesign of the usability and protective function for the detergent and milk cases, respectively. The results showed the significance of understanding the design consequences on the product life cycle by integrating the functional and environmental aspects.     

Cellulose nanoparticles synthesised from potato peel for the development of active packaging film for enhancement of shelf life of raw prawns (Penaeus monodon) during frozen storage

In this study, the cellulose nanoparticles (CNP) isolated from potato peel were used for reinforcement of polyvinyl alcohol (PVA)-based active packaging film. The above film was used to pack the raw prawns (Penaeus monodon) at −20 °C, and the colour change, protein content, TVB-N, TMA and microbial analysis were done at regular interval for prawns stored in CNP-PVA active packaging film. A significant difference was observed in the quality of prawns stored in potato CNP-PVA film compared with prawns packed and stored in polyethylene film. The newly designed active packaging with CNP and fennel seed oil enhanced the shelf life of prawns up to two months for both HOSO (head on shell on) prawn and PD (peeled and deveined) prawn. Hence, the study recommends the potato peel CNP-PVA film with fennel seed oil as better choice to extend the shelf life of the prawns during storage compared with polyethylene packaging.     

冷链条件下气调包装酱卤鸭肉制品货架期预测模型的建立

目的 以气调包装酱卤鸭肉制品为研究对象,在冷链温度范围内建立一套准确、高效的货架期预测模型。方法 利用选择性培养基测定不同温度下产品各微生物数量,确定4~25℃条件下产品优势腐败菌。对乳酸菌数量与感官评定值进行了回归分析确定最小腐败量Ns。分别采用修正的Gompertz方程和平方根方程建立一、二级模型,并通过预测值与实测值对比验证模型的可靠性。结果 确定了4~25℃条件下产品优势腐败菌为乳酸菌,最小腐败量Ns=6.14(lg(cfu /g))。一、二级模型拟合度均良好,三种温度下模型预测值与实际值间的差异均在30％左右,波动幅度在10%以内。结论 实现了对4~25℃内任何时间点产品剩余货架期的预测,为冷链条件下气调包装酱卤鸭肉制品品质的变化提供了理论指导。     

Synergistic effects of modified TiO2/multifunctionalized graphene oxide nanosheets as functional hybrid nanofiller in enhancing the interface compatibility of PLA/starch nanocomposites

The aim of this study is to investigate the synergistic effects of modified TiO₂/multifunctionalized graphene oxide nanosheets at different ratios on the interface compatibility between starch and poly(lactic acid) (PLA). To this end, silanylated nano-TiO₂ (MTiO₂, 1 and 2%) and alkylated maleic anhydride grafted graphene oxide (f-GO, 0.1, 0.2, and 0.4%) at different combinations are blended with the PLA-starch composites using solution blending technique. Then, the synergistic effects of MTiO₂ and f-GO on PLA/starch matrix are investigated in terms of the morphology, crystallinity, structural characterization, thermal stability, dynamic mechanical, and antiaging properties, and the related mechanisms. The Raman and Fourier transform infrared spectroscopy spectra verify the successful synthesis of the two modified nanofillers (f-GO and MTiO₂) and the formation of strong hydrogen bond within the PLA-starch nanocomposites. Due to the strong interfacial interaction and the synergistic effect from the combination of 1% MTiO₂ and 0.2% f-GO, obvious improvement was observed in PLA-starch versus other nanocomposites in terms of morphology, thermal stability, surface hydrophobicity, storage modulus, ultraviolet-shielding capacity, and aging-resistance. Furthermore, differential scanning calorimeter (DSC), isothermal crystallization kinetic, and X-ray diffraction analysis demonstrate that f-GO and the M-TiO₂ significantly synergize in enhancing the crystallization rate and crystallinity of PLA/starch matrix. These results provide novel insights for constructing high-performance nanocomposites and facilitate their applications in food packaging.     

Recycling of multilayer packaging using a reversible cross-linking adhesive

Plastic-based multilayer packaging has an important function on the packaging market, but is currently not recyclable as the polymer layers used are usually thermodynamically immiscible. This work therefore follows the approach to prepare separable multilayer packaging using a packaging adhesive modified with thermally unstable adducts, and proposes a corresponding recycling process. For this purpose, typical multilayer structures (polyethylene (PE)// polyethylene terephtalate (PET), PET//aluminum, and PE//aluminum) were prepared by curing furan-/maleimide-functionalized polyurethane (PU)-prepolymers with a three-functional cross-linking agent. Adhesions of up to over 3N per 15 mm test specimen were measured or substrate failures of PET films were observed. However, heating in dimethylsulfoxide, the retro-Diels–Alder reaction takes place and the cross-linked adhesive turns thermoplastic and dissolves in the solvent. Thus, the laminate separates and the pure PE, PET, and aluminum foils can be recovered without any PU residue.     

基于深度学习的视频跟踪研究进展综述

近年来深度学习迅猛发展，颠覆了语音识别、图像分类、文本理解等领域的算法设计思路。深度学习因其具备强大的特征提取能力，在图像识别领域的成绩尤为突出。然而深度学习与视频监控领域的结合并不多，由于深度模型具有多层网络结构，算法复杂度大，训练和更新模型时比较耗时，很难满足实时性要求。回顾了深度学习的发展史，介绍了最近10年来国内外深度学习主要模型，论述了基于深度学习的目标跟踪算法，指出了各算法的优缺点，最后对当前该领域存在的问题和发展前景进行了总结和展望。     

视频侦查中人像智能分析应用及算法优化

人像智能分析指的是对视频或录像中的人像进行结构化和可视化分析,对目标人物进行性别、年龄、发型等特征的智能识别,这项技术在视频侦查中有极高的应用价值。人像识别早期的算法是通过人工提取特征,通过学习低级视觉特征来针对不同属性进行分类学习,这种基于传统方法的模型表现常常不尽如人意。在计算机视觉领域,通过海量图像数据学习的神经网络比传统方法有更丰富的信息量和特征可以被提取。文章尝试通过深度学习技术训练神经网络模型对行人进行检测和识别,对于衣着不同的行人进行智能识别,具有更好的鲁棒性,提升了视频人像识别的准确率,拓展了人工智能技术在身份识别领域的应用。