1、Team # 1648 Page 1 of 21 第三届“认证杯”数学中国 数学建模国际赛 承 诺 书 我们仔细阅读了第三届“认证杯”数学中国数学建模国际赛的竞赛规则。 我们完全明白，在竞赛开始后参赛队员不能以任何方式（包括电话、电子邮件、网上咨询等）与队外的任何人（包括指导教师）研究、讨论与赛题有关的问题。 我们知道，抄袭别人的成果是违反竞赛规则的 , 如果引用别人的成果或其他公开的资料（包括网上查到的资料），必须按照规定的参考文献的表述方式在正文引用处和参考文献中明确列出。 我们郑重承诺，严格遵守竞赛规则，以保证竞赛的公正、公平性。如有违反竞赛规则 的行为，我们将受到严肃处理。 我们允许数
2、学中国网站 ()公布论文，以供网友之间学习交流，数学中国网站以非商业目的的论文交流不需要提前取得我们的同意。 我们的参赛队号为： 1648 我们选择的题目是： B 参赛队员 (签名 ) ： 队员 1: Qiang Wang 队员 2: Jian Ling Huang 队员 3： Zi Long Wang 参赛队教练员 (签名 )： 指导教师组 Team # 1648 Page 2 of 21 第三届“认证杯”数学中国 数学建模国际赛 编 号 专 用 页 参赛队伍的参赛队号： 1648 竞赛统一编号（由竞赛组委会送至评委团前编号）： 竞赛评阅编号（由竞赛评委团评阅前进行编号）： Team # 1
3、648 Page 3 of 21 Based on the fuzzy comprehensive evaluation and image segmentation research for camouflage pattern Abstract： Camouflage is mainly used in military camouflage cover aspects of performance, it is necessary to carry a specific test and evaluate for the performance of its hide and camou
4、flage, on the basis of the traditional single evaluation, the evaluation proposed a system, that would fuzzy comprehensive evaluation as a basis for the evaluation of a number of camouflage index and solves camouflage single evaluation method ,and the evaluation results of the lack of objectivity. A
5、ccording to the purposes and object of the evaluation system, this method comprehensive selection of the camouflage color, shape, and texture(such as the environment belong, regional distribution, regional context, size of spot, size of the camouflage material, brightness contrast, and edge shape) f
6、eatures as the evaluation system index. Use of the Single-level fuzzy evaluation method nbbbRAB ,.,21 and Multi-level fuzzy evaluation method *B A R take comprehensive analyze and calculation on the factor that would influent the camouflage effect. In the embodiment of the process of evaluation adop
7、t the fuzzy method and the precise method camouflage effect on various practical investigation materials, various experimental data and research results and other indicators to quantify application. Using the weight method to determine the impact of each index value weights and work out each of fact
8、or evaluation rank at the end. According to the evaluation grade weighting coefficients could distinguish the specific camouflage in the particular mode effect. On this basic for the judge of the camouflage effect evaluation application image region segmentation method to take turns have a similarit
9、y calculation Kk IIII lkSKS jiji 1 , ,m a x1 on original picture and camouflage picture. I n addition to, under the different original feature we obtain a calculation formula ),(_),(_),(_),( lkSSlkSHlkSIwlkSJPJPJPikJP IIIIIIIII , then use Matlab on the region segmentation similarity solution princip
10、les to calculation the region similarity under the different region feature. There by improving the accuracy and feasibility of the calculation. Team # 1648 Page 4 of 21 Finally, using the above methods to take quantification indicators and get the ideal of design camouflage patterns, work out ourse
11、lves camouflage patterns, camouflage pattern designed close to the expected idea after correcting by Photoshop and other software, after test, camouflage patterns effect mainly reach to the request of above-mentioned evaluations. Key words: Fuzzy comprehensive evaluation method Region segmentation S
12、imilarity Camouflage Team # 1648 Page 5 of 21 Contents I. Problem restatement . 6 II. Problem analysis. 6 III. Symbol description . 7 IV. Model assumption . 7 V. Establishment and Solution of the model . 8 5.1Established camouflage evaluation model . 8 5.2 Determining the weight subset of evaluation
13、 factors. 9 5.3 Fuzzy Evaluation Model . 9 5.4 Solving weights. 12 5.5 Image region segmentation and regional characteristics . 14 5.6 Design pattern . 17 5.6.1 Design and evaluate patterns are as follow . 17 5.6.2 Advertising Design. 17 VI. Evaluation and extension . 18 VII. References . 19 VIII. A
14、ppendix . 19 Team # 1648 Page 6 of 21 I. Problem restatement Wearing grey camouflage clothes in the barren desert, less likely to be attacked by enemy. In the vast expanse of snowy forest, the soldiers wearing white camouflage, military operations, so that the enemy is difficult to find in the vast
15、expanse of snowy in order to protect the victory in battle. In the vast expanse of snowy forest, the soldiers wearing white camouflage, military operations, so that the enemy is difficult to find in the vast expanse of snow ,so that it can protect the victory in battle. Since human civilization born
16、, people have been using some form of camouflage. Military developed today. From clothing to weaponry, military camouflage has become a high technology content in systematic technique. It is a important content and essential for modern military. Camouflage workers also known as ground war magician.
17、And color plays a vital role in military camouflage. How to build a model to analyze the effect of a variety of camouflage patterns when visibility is low , Another effect of the development of a quantitative evaluation methods and camouflage design makes an ideal camouflage pattern can be applied t
18、o most environments， a quantitative evaluation method development and design of a camouflage effect applies to most ideal environment camouflage pattern, in addition , how to find a quantitative evaluation method and design a camouflage pattern applies to most environment . II. Problem analysis When
19、 military targets camouflage, camouflage effect how very important goal disguised whether the same or similar background to determine the target is the probability of enemy reconnaissance, so evaluation of their camouflage effect is extremely important. The traditional evaluation method is by a prof
20、essional target person or photograph observation, judge camouflage effect, subjective factors affecting the subject of relatively large, in order to camouflage the effect of the objective quantitative description of a method discussed herein used to evaluate the similarity image camouflage effect th
21、e image characteristic generally can color, shape and texture characteristics described, that is, whether the similarity is performed by the similar characteristics, that is, the comparison between the characteristics of the image as a comparison image to judge the similarity between the images. Tea
22、m # 1648 Page 7 of 21 III. Symbol description Symbol Explanation B Rank U Evaluation elements iU A subset of the elements CTPf Regional texture features ,higstogramf Color histogram area shapef Area shape characteristic kld Two regional integrated feature of the Euclidean distance Skld Region shape
23、features two Euclidean distance IV. Model assumption Hypothesis one: Suppose weight vector and fuzzy matrix size in most cases are matched, there are no super blur; Hypothesis two: Suppose ignored region segmentation fuzzy comprehensive evaluation weighting factor the impact; Hypothesis three: Suppo
24、se when the similarity is determined with reference to two pictures of the same conditions, ignore the effects of small factors; Team # 1648 Page 8 of 21 V. Establishment and Solution of the model 5.1Established camouflage evaluation model In modern warfare, in order to deal with high-tech weaponry
25、enemy observation, reconnaissance and attack, defending their own strategic and tactical objectives and effective strength, improve survival and combat troops, a large number of camouflage technology implementation is essential. Camouflage its fast convergence, low cost, easy construction and other
26、advantages are widely used in military camouflage goal is a common means of military reconnaissance and weapon systems against attack, the merits of its relationship to combat camouflage effect can be effectively preservation. Whether in peacetime or wartime, camouflage for the protection of war mat
27、erials to improve combat effectiveness and battlefield survivability has a very important role. Thus that when military targets camouflage, camouflage effect how very important goal disguised whether the same or similar background to determine the target is the probability of enemy reconnaissance, s
28、o evaluation of their camouflage effect is extremely important. This article will disguise camouflage rating factors are divided into: color, texture, shape features. Single factor is: Quantity environmental, geographical distribution, size, regional background, spot size, camouflage material, which
29、 the season, weather conditions, brightness and contrast, edge shape, spatial distribution, this article only correlation analysis of the main factors Evaluation elements set is: , 321 uuuU where each single subset of features )3,2,1( iUi area as follows , 1312111 uuuU , 25242322212 uuuuuU , 3332313
30、 uuuU Comments collection to determine According to the actual needs assessment decisions will judge grading standards are divided into “good“, “better“, “fair“, “poor“ and “bad five levels. That comment set as follows: , 54321 VVVVVV = good, better ,fair, poor, bad Team # 1648 Page 9 of 21 5.2 Dete
31、rmining the weight subset of evaluation factors In the comprehensive evaluation index system above camouflage, since the lower the relative importance of each indicator on the top of some indicators are not the same, some of the indicators of the degree of influence that is greater than or more than
32、 other indicators. Therefore, in order to lower the relative importance of each indicator to measure the upper indicators, evaluation of the need to determine the right weight coefficient. Common determine the weight coefficient method are: (1) subjective experience to judge law; (2) the expert inve
33、stigation or expert consultation method; (3) the group judgment of experts to brainstorm voting method; (4) analytic hierarchy process (ie, AHP method). To ensure objectivity, fairness and scientific coefficients determined weight, which often several methods described above can be used in combinati
34、on. Following the use of subjective experience and expert judgment Combination consulted to determine levels of evaluation factor index weight coefficient subset. Each sub-centralization weight (a weight) as follows： , 321 aaaA 34.0,28.0,0 3 8A ： Each subset of all elements of weights (two weights)
35、is: , 1312111 aaaA , 25242322212 aaaaaA , 3332313 aaaA 29.033.038.01 ，A 13.014.020.028.025.02 ，A 32.033.035.03 ，A 5.3 Fuzzy Evaluation Model Fuzzy comprehensive evaluation method is an application of fuzzy mathematics to analyze and evaluate with the “fuzziness“ of things, systems analysis. It is a
36、kind of fuzzy inference based qualitative and quantitative precision and non-precision unity of analysis and evaluation methods. Since this method is difficult to use in dealing with a variety of precise mathematical description of complex systems problems exhibited unique advantages, in recent year
37、s has been very widely used in many subject areas. Team # 1648 Page 10 of 21 A : single-level fuzzy comprehensive evaluation model: Given two finite domain： , 21 muuuU （ 1） , 21 mVVVV （ 2） (1) Where, U represents a collection of all the evaluation factors consisting of; (2) where, V represents a col
38、lection of reviews of all grades composed. If one looks at the first evaluation factor ),2,1( mii , the single factor evaluation result is , 11211 mrrrR : Judgment is a judgment decision matrix of m factors: 1 11 12 12 21 22 212nnm m m m nR r r rR r r rRR r r r （ 3） A fuzzy relation is U toV . If th
39、e weights assigned to each evaluation factors are1, 2 ,mA a a a (apparently, is one fuzzy subset of the domain U, and10 1, 1miiiaa ) the conversion of synthesis of fuzzy arithmetic, you can get a fuzzy subset of the domain V, ie comprehensive Evaluation of the results: nbbbRAB ,.,21 (4) B. multi-level fuzzy comprehensive evaluation model a. Set of evaluation factorsU , according to a property, which is divided into m subsets, so that they meet: 1()miiijUUU U i j （ 5） Thus, we get a set of the second-stage evaluation factors: , 21 mUUUU （ 6）
