Please wait a minute...
浙江大学学报(工学版)  2021, Vol. 55 Issue (3): 419-429    DOI: 10.3785/j.issn.1008-973X.2021.03.001
机械工程     
面向汽车主客观审美评价的不确定性推理模型
周爱民(),刘宏斌,张书涛*(),欧阳晋焱
兰州理工大学 设计艺术学院,甘肃 兰州 730050
Uncertainty reasoning model of subjective and objective aesthetic evaluation of car
Ai-min ZHOU(),Hong-bin LIU,Shu-tao ZHANG*(),Jin-yan OUYANG
School of Design Art, Lanzhou University of Technology, Lanzhou 730050, China
 全文: PDF(898 KB)   HTML
摘要:

为了克服主、客观单一评价方法的片面性和局限性,从不确定信息推理的角度,提出结合主、客观评价证据源进行信息融合的汽车前脸形态审美评价模型. 依据人类视觉审美原理与计算美学方法,构建汽车前脸美度指标体系及指标计算公式. 运用两极递进排序法调查,获得审美评价值;依据用户的审美偏好进行审美群体分类,运用灰关联分析法计算指标权重,得到各群体主观评价的证据. 运用熵值法和相关性定权法计算指标权重,得到2种客观审美评价的证据. 基于Dempster-Shafer证据理论对主、客观证据信息的可信度进行融合推理,实现汽车前脸形态审美评价. 经实验验证,该方法能较好地平衡审美主体的主观意愿与客体的客观反映,具有较好的合理性与可靠性.

关键词: 汽车前脸不确定性推理形态审美评价Dempster-Shafer证据理论    
Abstract:

From the perspective of uncertain information reasoning, a aesthetic evaluation model of car front face form was proposed by combining subjective and objective evaluation evidence sources for information fusion, in order to overcome the one-sidedness and limitation of the single method of subjective evaluation and objective evaluation. Firstly, according to the principle of human visual aesthetics and the computational aesthetics method, the system of aesthetic measure and index calculation formula was constructed. Secondly, a questionnaire survey was carried out by using the two-pole progressive sorting method to obtain the aesthetic evaluation values. According to aesthetic preferences, users were classified into aesthetic groups. Using the grey relation analysis method, the weight of each index was calculated, and the evidence of subjective evaluation of each aesthetic group was acquired. Thirdly, the weight of each index was calculated using the entropy method and the CRITIC method, and two evidence sources of objective aesthetic evaluation were obtained. Finally, based on Dempster-Shafer evidence theory, the credibility of the subjective/objective evidence information was integrated and reasoned to evaluate the car front face’s aesthetic. Experimental results show that the method can balance the subjective will of the aesthetic subject and the objective reflection of the object, and has good rationality and reliability.

Key words: car front face    uncertainty reasoning    form    aesthetic evaluation    Dempster-Shafer evidence theory
收稿日期: 2020-01-27 出版日期: 2021-04-25
CLC:  TH 166  
基金资助: 国家自然科学基金资助项目(51705226);甘肃省自然科学基金资助项目(2017gs10786)
通讯作者: 张书涛     E-mail: 51289547@qq.com;364725955@qq.com
作者简介: 周爱民(1978—),男,副教授,从事产品形态智能设计研究. orcid.org/0000-0002-3994-9040. E-mail: 51289547@qq.com
服务  
把本文推荐给朋友
加入引用管理器
E-mail Alert
作者相关文章  
周爱民
刘宏斌
张书涛
欧阳晋焱

引用本文:

周爱民,刘宏斌,张书涛,欧阳晋焱. 面向汽车主客观审美评价的不确定性推理模型[J]. 浙江大学学报(工学版), 2021, 55(3): 419-429.

Ai-min ZHOU,Hong-bin LIU,Shu-tao ZHANG,Jin-yan OUYANG. Uncertainty reasoning model of subjective and objective aesthetic evaluation of car. Journal of ZheJiang University (Engineering Science), 2021, 55(3): 419-429.

链接本文:

http://www.zjujournals.com/eng/CN/10.3785/j.issn.1008-973X.2021.03.001        http://www.zjujournals.com/eng/CN/Y2021/V55/I3/419

图 1  人类视觉认知系统模型
图 2  美度指标计算坐标图
类型 布局特征 示例
1 6个元素相互独立 铃木雨燕
2 上下进气格栅紧密相连,
上下车灯相互独立
标致308
3 上车灯与上进气格栅紧密相连,
下车灯与下格栅相互独立
本田思域
4 下车灯与下进气格栅紧密相连,
上车灯与上进气格栅相互独立
别克君越
5 上车灯与上下进气格栅紧密相连,
下车灯相互独立
日产骐达
6 上车灯与上进气格栅紧密相连,
下车灯与下进气格栅紧密相连
大众辉腾
表 1  汽车前脸布局特征和示例
样本 美度指标值
X1 X2 X3 X4 X5 X6 X7 X8 X9 X10 X11 X12 X13
1 0.859 0.962 0.805 0.731 0.333 0.333 0.548 0.588 0.769 0.444 0.333 0.444 0.222
3 0.785 0.942 0.814 0.754 0.208 0.333 0.533 0.673 0.623 0.444 0.444 0.333 0.333
4 0.803 0.951 0.762 0.679 0.271 0.333 0.571 0.707 0.584 0.333 0.333 0.222 0.222
5 0.817 0.953 0.787 0.753 0.250 0.333 0.566 0.618 0.456 0.444 0.444 0.000 0.333
6 0.807 0.956 0.792 0.770 0.229 0.333 0.593 0.704 0.614 0.333 0.333 0.222 0.222
7 0.886 0.976 0.725 0.837 0.200 0.375 0.599 0.806 0.481 0.250 0.375 0.000 0.250
8 0.804 0.944 0.799 0.684 0.292 0.333 0.513 0.657 0.535 0.222 0.333 0.222 0.333
9 0.792 0.951 0.789 0.680 0.354 0.333 0.486 0.668 0.478 0.333 0.444 0.000 0.222
10 0.792 0.952 0.766 0.686 0.250 0.286 0.507 0.734 0.562 0.286 0.286 0.286 0.143
11 0.775 0.946 0.801 0.644 0.250 0.286 0.516 0.631 0.533 0.500 0.429 0.286 0.143
12 0.757 0.942 0.786 0.628 0.188 0.286 0.491 0.725 0.462 0.429 0.286 0.000 0.143
13 0.842 0.968 0.766 0.701 0.225 0.250 0.519 0.662 0.411 0.313 0.250 0.125 0.313
14 0.764 0.939 0.800 0.735 0.219 0.286 0.564 0.568 0.440 0.286 0.286 0.286 0.286
15 0.842 0.958 0.766 0.676 0.208 0.286 0.483 0.647 0.526 0.500 0.286 0.286 0.143
表 2  15个汽车前脸样本的各美度指标值
图 3  15个汽车前脸图
样本 美感调查值
第1类 第2类 第3类 第4类 第5类 第6类
1 0.456 0.844 0.603 0.547 0.936 0.347
2 0.478 0.711 0.552 0.500 0.491 0.000
3 0.467 0.412 0.466 0.519 0.164 0.939
4 0.867 0.853 0.759 0.745 0.573 0.224
5 0.000 0.758 0.983 0.670 0.555 0.592
6 0.444 0.403 0.500 0.443 0.991 0.224
7 0.300 0.991 0.724 0.340 0.209 0.796
8 0.222 0.678 0.138 0.245 1.000 0.776
9 0.644 1.000 0.362 0.377 0.982 0.020
10 0.522 0.118 0.069 0.858 0.536 0.347
11 0.433 0.118 1.000 0.792 0.536 0.449
12 0.878 0.664 0.000 0.670 0.327 0.755
13 1.000 0.474 0.569 0.358 0.482 0.204
14 0.300 0.507 0.293 1.000 0.000 0.061
15 0.822 0.000 0.483 0.000 0.945 1.000
表 3  6个审美群体的美感调查值
群体 ω1 ω2 ω3 ω4 ω5 ω6 ω7 ω8 ω9 ω10 ω11 ω12 ω13
第1类 0.083 0.091 0.069 0.057 0.080 0.082 0.077 0.085 0.084 0.076 0.070 0.078 0.067
第2类 0.074 0.067 0.083 0.086 0.074 0.099 0.092 0.069 0.060 0.085 0.073 0.063 0.076
第3类 0.089 0.075 0.061 0.084 0.068 0.080 0.086 0.069 0.074 0.082 0.075 0.073 0.084
第4类 0.072 0.064 0.082 0.071 0.071 0.071 0.096 0.087 0.076 0.086 0.067 0.082 0.077
第5类 0.089 0.109 0.082 0.071 0.102 0.076 0.059 0.078 0.083 0.071 0.060 0.069 0.051
第6类 0.080 0.067 0.083 0.076 0.083 0.080 0.076 0.071 0.073 0.084 0.080 0.069 0.076
表 4  6个审美群体的各美度指标权重
方法 W1 W2 W3 W4 W5 W6 W7 W8 W9 W10 W11 W12 W13
熵值法 0.041 0.027 0.056 0.051 0.109 0.109 0.076 0.057 0.070 0.102 0.103 0.095 0.103
CRITIC法 0.062 0.068 0.078 0.062 0.077 0.065 0.089 0.068 0.050 0.095 0.098 0.077 0.111
表 5  2种客观评价的各美度指标权重
图 4  5种评价方法的评价结果比较
1 罗仕鉴, 李文杰, 傅业焘 消费者偏好驱动的SUV产品族侧面外形基因设计[J]. 机械工程学报, 2016, 52 (2): 173- 181
LUO Shi-jian, LI Wen-jie, FU Ye-tao Consumer preference driven SUV product family profile gene design[J]. Journal of Mechanical Engineering, 2016, 52 (2): 173- 181
doi: 10.3901/JME.2016.02.173
2 LUGO J E, SCHMIEDELER J P, BATILL S M, et al Relationship between product aesthetic subject preference and quantified Gestalt principles in automobile wheel rims[J]. Journal of Mechanical Design, 2016, 138 (5): 051101
doi: 10.1115/1.4032775
3 ORSBORN S, CAGAN J, BOAT W P Quantifying aesthetic form preference in a utility function[J]. Journal of Mechanical Design, 2009, 131 (6): 397- 407
4 TANG C Y, FUNG K Y, LEE E W M, et al Product form design using customer perception evaluation by a combined superellipse fitting and ANN approach[J]. Advanced Engineering Informatics, 2013, 27 (3): 386- 394
doi: 10.1016/j.aei.2013.03.006
5 CHEN H Y, CHANG H C Consumers’ perception-oriented product form design using multiple regression analysis and backpropagation neural network[J]. Artificial Intelligence for Engineering Design, Analysis and Manufacturing, 2016, 30 (1): 64- 77
doi: 10.1017/S0890060415000165
6 BIRKHOFF G D. Aesthetic measure [M]. Cambridge: Harvard University Press, 1933: 4.
7 VALENCIA R A, LUGO J E Part-worth utilities of Gestalt principles for product aesthetics: a case study of a bottle silhouette[J]. Journal of Mechanical Design, 2016, 138 (8): 081102
doi: 10.1115/1.4033664
8 ZHOU L, XUE C Q, TOMIMATSU K Research of interface composition design optimization based on visual balance[J]. Advances in Intelligent Systems and Computing, 2014, 279 (1): 483- 493
9 NGO D C L, TEO L S, BYME J G Modelling interface aesthetics[J]. Information Sciences, 2003, 152: 25- 46
doi: 10.1016/S0020-0255(02)00404-8
10 WANNARUMON S, BOHEZ E L J A new aesthetic evolutionary approach for jewelry design[J]. Computer-Aided Design and Applications, 2006, 3 (1-4): 385- 394
doi: 10.1080/16864360.2006.10738477
11 LO C H, KO Y C, HSIAO S W A study that applies aesthetic theory and genetic algorithms to product form optimization[J]. Advanced Engineering Informatics, 2015, 29 (3): 662- 679
doi: 10.1016/j.aei.2015.06.004
12 曹卫华, 杜楠, 安剑奇, 等 基于主客观证据融合的高炉悬料预测方法[J]. 北京科技大学学报, 2014, 36 (4): 506- 514
CAO Wei-hua, DU Nan, AN Jian-qi, et al Prediction method of blast furnace hanging based on fusion of subjective and objective evidences[J]. Journal of University of Science and Technology Beijing, 2014, 36 (4): 506- 514
13 谭浩, 赵江洪, 赵丹华, 等 汽车造型特征定量模型构建与应用[J]. 湖南大学学报: 自然科学版, 2009, 36 (11): 27- 31
TAN Hao, ZHAO Jiang-hong, ZHAO Dan-hua, et al Construction and application of the quantitative model of automobile form features[J]. Journal of Hunan University: Natual Science Edition, 2009, 36 (11): 27- 31
14 梁宁建. 当代认知心理学[M]. 上海: 上海教育出版社, 2014: 6.
15 周爱民, 苏建宁, 阎树田, 等 产品形态审美综合评价的非线性信息动力学模型[J]. 机械工程学报, 2018, 54 (15): 150- 159
ZHOU Ai-min, SU Jian-ning, YAN Shu-tian, et al Nonlinear information dynamics model of synthetic evaluation on product form aesthetic[J]. Journal of Mechanical Engineering, 2018, 54 (15): 150- 159
doi: 10.3901/JME.2018.15.150
16 阿恩海姆. 艺术与视知觉[M]. 成都: 四川人民出版社, 2019: 54.
17 CAPPADONA F, GOUSSARD J, SUTRA L, et al. Fiores ii: a quantitative approach of aesthetic notions [C]// Collaborative Design MICAD Conf. Pairs: [s.n.] , 2003: 1-9.
18 王硕楠, 余宏明, 刘运涛 基于灰关联度的地质灾害易损性区划研究[J]. 安全与环境工程, 2011, 18 (2): 10- 13
WANG Shuo-nan, YU Hong-ming, LIU Yun-tao Study on the regional vulnerability of geological hazard based on grey relative degree[J]. Safety and Environmental Engineering, 2011, 18 (2): 10- 13
doi: 10.3969/j.issn.1671-1556.2011.02.003
19 张红涛, 毛罕平 四种客观权重确定方法在粮虫可拓分类中的应用比较[J]. 农业工程学报, 2009, 25 (1): 132- 136
ZHANG Hong-tao, MAO Han-ping Comparison of four methods for deciding objective weights of features for classifying stored-grain insects based on extension theory[J]. Transactions of the Chinese Society of Agricultural Engineering, 2009, 25 (1): 132- 136
20 赵丽, 朱永明, 付梅臣, 等 主成分分析法和熵值法在农村居民点集约利用评价中的比较[J]. 农业工程学报, 2012, 28 (7): 235- 242
ZHAO Li, ZHU Yong-ming, FU Mei-chen, et al Comparative study on intensive use of rural residential land based on principal component analysis and entropy[J]. Transactions of the Chinese Society of Agricultural Engineering, 2012, 28 (7): 235- 242
doi: 10.3969/j.issn.1002-6819.2012.07.039
21 DIAKOULAKI D, MAVROTAS G, PAPAYANNAKIS L Determining objective weights in multiple criteria problems: the critic method[J]. Computers and Operations Research, 1995, 22 (7): 763- 770
doi: 10.1016/0305-0548(94)00059-H
22 谢彦蓉, 李孜军, 徐志国 基于CRITIC法与TOPSIS法的硫化矿自燃倾向性评定[J]. 安全与环境学报, 2014, 14 (1): 122- 125
XIE Yan-rong, LI Zi-jun, XU Zhi-guo Evaluation on spontaneous combustion trend of sulfide ores based on the method of CRITIC and TOPSIS testing method[J]. Journal of Safety and Environment, 2014, 14 (1): 122- 125
23 DEMPSTER A P Upper and lower probabilities induced by a multi-valued mapping[J]. The Annals of Mathematical Statistics, 1967, 38 (2): 325- 339
doi: 10.1214/aoms/1177698950
24 LEUNG Y, LI R, JI N Application of extended Dempster-Shafer theory of evidence in accident probability estimation for dangerous goods transportation[J]. Journal of Geographical Systems, 2017, 19 (3): 249- 271
doi: 10.1007/s10109-017-0253-2
25 由东媛, 曹梦龙, 姜凯 D-S证据理论中冲突证据的改进方法研究[J]. 电子测量技术, 2018, 41 (23): 35- 39
YOU Dong-yuan, CAO Meng-long, JIANG Kai Research on improved method of conflict evidence in D-S evidence theory[J]. Electronic Measurement Technology, 2018, 41 (23): 35- 39
26 汪振双, 周梅 基于价值工程原理的煤矸石混凝土配合比设计方案选择[J]. 数学的实践与认识, 2015, 45 (7): 126- 132
WANG Zhen-shuang, ZHOU Mei Study on mix proportion of coal gangue concrete on value engineering theory[J]. Mathematics in Practice and Theory, 2015, 45 (7): 126- 132
27 朱英菊, 刘红丽, 陈长松 信息安全管理有效性的测量研究[J]. 情报杂志, 2010, 29 (1): 73- 76
ZHU Ying-jiu, LIU Hong-li, CHEN Chang-song Research of measuring information security management effectiveness[J]. Journal of Intelligence, 2010, 29 (1): 73- 76
doi: 10.3969/j.issn.1002-1965.2010.01.016
28 毛国柱, 侯长胜, 柴立和, 等 基于最大流原理的草型与藻型湖泊富营养化驱动因子识别[J]. 环境工程学报, 2016, 10 (2): 768- 774
MAO Guo-zhu, HOU Chang-sheng, CHAI Li-he, et al Identification of eutrophication driving factors of grass-type lakes and algae-type lakes based on maximum flux principle[J]. Chinese Journal of Environmental Engineering, 2016, 10 (2): 768- 774
doi: 10.12030/j.cjee.20160240
29 李少波, 全华凤, 胡建军, 等 基于在线评论数据驱动的产品感性评价方法[J]. 计算机集成制造系统, 2018, 24 (3): 752- 762
LI Shao-bo, QUAN Hua-feng, HU Jian-jun, et al Perceptual evaluation method of products based on online reviews data driven[J]. Computer Integrated Manufacturing Systems, 2018, 24 (3): 752- 762
30 宋光兴, 杨德礼 基于决策者偏好及赋权法一致性的组合赋权法[J]. 系统工程与电子技术, 2004, 26 (9): 1226- 1230
SONG Guang-xing, YANG De-li Combination weighting approach based on the decision-maker’s preference and consistency of weighting methods[J]. Systems Engineering and Electronics, 2004, 26 (9): 1226- 1230
doi: 10.3321/j.issn:1001-506X.2004.09.020
31 倪明, 高晓萍, 单渊达 证据理论在中期负荷预测中的应用[J]. 中国电机工程学报, 1997, 17 (3): 56- 60
NI Ming, GAO Xiao-ping, SHAN Yuan-da Application of evidential theory in middle-term load forecasting[J]. Proceedings of the CSEE, 1997, 17 (3): 56- 60
32 唐忠, 李文强, 李彦 含广义三角模糊数的机械结构可靠度计算方法[J]. 系统工程理论与实践, 2018, 38 (8): 2155- 2167
TANG Zhong, LI Wen-qiang, LI Yan Reliability calculation method for mechanical structures with generalized triangular fuzzy number[J]. Systems Engineering—Theory and Practice, 2018, 38 (8): 2155- 2167
doi: 10.12011/1000-6788(2018)08-2155-13
33 韩德强, 杨艺, 韩崇昭 DS证据理论研究进展及相关问题探讨[J]. 控制与决策, 2014, 29 (1): 1- 11
HAN De-qiang, YANG Yi, HAN Chong-zhao Advances in DS evidence theory and related discussions[J]. Control and Decision, 2014, 29 (1): 1- 11
[1] 苗发盛,吴益平,李麟玮,廖康,薛阳. 基于Boosting-决策树C5.0的岩体结构面粗糙度预测[J]. 浙江大学学报(工学版), 2021, 55(3): 483-490.
[2] 冉树浩,胡玉龙,杨元维,高贤君,李熙,陈明珠. 基于样本形态变换的高分遥感影像建筑物提取[J]. 浙江大学学报(工学版), 2020, 54(5): 996-1006.
[3] 刘凯,吉小军,赵忠华,曹一文,杨剑,庞晓锋. 用于轿车PEPS系统的双终端差分改进识别算法[J]. 浙江大学学报(工学版), 2020, 54(10): 1892-1898.
[4] 卓新, 张沈斌. 直杆叠积螺旋空间结构的找形分析方法[J]. 浙江大学学报(工学版), 2018, 52(3): 413-419.
[5] 余建波, 李传锋, 吕靖香. 轴承故障信号的平均组合差值形态滤波分析[J]. 浙江大学学报(工学版), 2018, 52(10): 1845-1853.
[6] 葛丹东, 童磊, 吴宁, 华晨, 杜佳. 乡村道路形态参数化解析与重构方法[J]. 浙江大学学报(工学版), 2017, 51(2): 279-286.
[7] 侯霞丽,李晓东,陈彤,陆胜勇,纪莎莎,任咏. 垃圾焚烧飞灰中主要元素的深度分布及形态[J]. 浙江大学学报(工学版), 2015, 49(5): 930-937.
[8] 夏军强, 宗全利, 邓珊珊, 许全喜, 张翼. 三峡工程运用后荆江河段平滩河槽形态调整特点[J]. 浙江大学学报(工学版), 2015, 49(2): 238-245.
[9] 孙志林, 杨仲韬, 高运, 许丹, 胡世祥. 长江分汊河口水力几何形态[J]. 浙江大学学报(工学版), 2014, 48(12): 2266-2270.
[10] 陈越超, 周晓军, 杨辰龙, 李钊. L型CFRP构件R区微观形态及孔隙特征[J]. 浙江大学学报(工学版), 2014, 48(10): 1775-1880.
[11] 周水琴, 应义斌, 商德胜. 基于形态学的香梨褐变核磁共振成像无损检测[J]. J4, 2012, 46(12): 2141-2145.
[12] 刘曦泽, 祁国宁, 傅建中, 樊蓓蓓, 许静. 集成形态学矩阵与冲突解决原理的设计过程模型[J]. J4, 2012, 46(12): 2243-2251.
[13] 陈卫东, 李昕, 刘俊, 郝耀耀, 廖玉玺, 苏煜, 张韶岷, 郑筱祥. 基于数学形态学的眼电信号识别及其应用[J]. J4, 2011, 45(4): 644-649.
[14] 何胜, 周劲松, 朱燕群, 骆仲泱, 倪明江, 岑可法. 钒系SCR催化剂对汞形态转化的影响[J]. J4, 2010, 44(9): 1773-1780.
[15] 张文斌, 周晓军, 沈路, 李俊生, 杨先勇, 林勇. 基于形态小波的转子轴心轨迹提纯[J]. J4, 2010, 44(8): 1449-1453.