Journal of Intelligent & Fuzzy Systems - Volume 40, issue 3

Authors: Tang, Fei

Article Type: Research Article

Abstract: To improve the optimization efficiency of the intelligent bionic optimization algorithm, this paper proposes intelligent bionic optimization algorithm based on the growth characteristics of tree branches. Firstly, the growth organ of the tree is mapped into the coding of the tree growth algorithm (intelligent bionic optimization algorithm). Secondly, the entire tree, that is the growing tree, is formed by selecting the individual that grows fast to generate the next level of shoot population. Lastly, if the growing tree reaches a certain level, the individual coding of the shoots is added to enhance the searching ability of the individuals of current …generation in the growth tree growth space, so that the algorithm approaches the optimal solution. The experimental results were compared with the optimization results of the genetic algorithm and the ant colony algorithm using the classic optimization function and showed that this algorithm has fewer iterations, a faster convergence speed, higher precision, and a better optimization ability than the genetic algorithm and the ant colony algorithm. Show more

Keywords: Individual coding, branch population, genetic algorithm, tree growth algorithm

DOI: 10.3233/JIFS-190487

Citation: Journal of Intelligent & Fuzzy Systems, vol. 40, no. 3, pp. 3821-3829, 2021

Authors: Lin, Jiang | Jianjun, Zhu

Article Type: Research Article

Abstract: As a new business form of innovation and development, new R&D institutions are characterized by their focus on regional and industrial major technical needs, diversified investment subjects, diversified construction models, and marketization of operation mechanism. Their performance evaluation faces new problems and challenges. This paper proposes a new dynamic grey target evaluation model of R&D institutions’ performance in regard to four evaluation indexes, three reference points, and four stages. Aiming at resolving the multi-attribute dynamic decision problem with the attribute value being an interval grey number and the decision-maker’s weight information unknown, we propose the use of the close degree …of grey incidence method to determine the index weight. Our approach revolves around three reference points: peers, development, and expectations. Value matrices of the three reference points are expressed according to the Cumulative Prospect Theory, which also determines the distance from the center of the grey target. Based on the Orness measure, we establish a multi-stage weight optimization model to calculate the stage weight and the comprehensive cumulative prospect value of each agency. Finally, we verify the validity and practicability of our method with the use of parameter sensitivity analysis, a comparison with other methods, and a case study. Show more

Keywords: Three reference points, new R&D institutions, performance, dynamic grey target, evaluation

DOI: 10.3233/JIFS-190602

Citation: Journal of Intelligent & Fuzzy Systems, vol. 40, no. 3, pp. 3831-3847, 2021

Authors: Yousaf, Waqas | Umar, Arif | Shirazi, Syed Hamad | Khan, Zakir | Razzak, Imran | Zaka, Mubina

Article Type: Research Article

Abstract: Automatic logo detection and recognition is significantly growing due to the increasing requirements of intelligent documents analysis and retrieval. The main problem to logo detection is intra-class variation, which is generated by the variation in image quality and degradation. The problem of misclassification also occurs while having tiny logo in large image with other objects. To address this problem, Patch-CNN is proposed for logo recognition which uses small patches of logos for training to solve the problem of misclassification. The classification is accomplished by dividing the logo images into small patches and threshold is applied to drop no logo area …according to ground truth. The architectures of AlexNet and ResNet are also used for logo detection. We propose a segmentation free architecture for the logo detection and recognition. In literature, the concept of region proposal generation is used to solve logo detection, but these techniques suffer in case of tiny logos. Proposed CNN is especially designed for extracting the detailed features from logo patches. So far, the technique has attained accuracy equals to 0.9901 with acceptable training and testing loss on the dataset used in this work. Show more

Keywords: Logo detection, logo recognition, deep learning, AlexNet, ResNet, CNN

DOI: 10.3233/JIFS-190660

Citation: Journal of Intelligent & Fuzzy Systems, vol. 40, no. 3, pp. 3849-3862, 2021

Authors: Dabighi, Korosh | Nazari, Akbar | Saryazdi, Saeid

Article Type: Research Article

Abstract: Nowadays, Canny edge detector is considered to be one of the best edge detection approaches for the images with step form. Various overgeneralized versions of these edge detectors have been offered up to now, e.g. Saryazdi edge detector. This paper proposes a new discrete version of edge detection which is obtained from Shen-Castan and Saryazdi filters by using bilinear transformation. Different experimentations are conducted to decide the suitable parameters of the proposed edge detector and to examine its validity. To evaluate the strength of the proposed model, the results are compared to Canny, Sobel, Prewitt, LOG and Saryazdi methods. Finally, …by calculation of mean square error (MSE) and peak signal-to-noise ratio (PSNR), the value of PSNR is always equal to or greater than the PSNR value of suggested methods. Moreover, by calculation of Baddeley’s error metric (BEM) on ten test images from the Berkeley Segmentation DataSet (BSDS), we show that the proposed method outperforms the other methods. Therefore, visual and quantitative comparison shows the efficiency and strength of proposed method. Show more

Keywords: Edge detection, Laplace operator, impulse response invariance, bilinear transformation

DOI: 10.3233/JIFS-191229

Citation: Journal of Intelligent & Fuzzy Systems, vol. 40, no. 3, pp. 3863-3874, 2021

Authors: Jian, Xianzhong | Wang, Xutao

Article Type: Research Article

Abstract: The existing methods for classification of power quality disturbance signals (PQDs) have the problems that the process of signal feature selection is tedious and imprecise, the accuracy of classification has no guiding significance for feature extraction, and lack of adequate labelled training data. To solve these problems, this paper proposes a new semi-supervised method for classification of PQDs based on generative adversarial network (GAN). Firstly, a GAN model is designed which we call it PQDGAN. After the unsupervised pre-training with unlabeled training data, the trained discriminator is extracted alone and conduct supervised training with a small amount of labelled training …data. Finally, the discriminator became a classifier with high accuracy. This model can achieve the step of feature extraction and selection efficiently. In addition, only a small amount of labelled training data is used, which greatly reduces the dependence of classification model on labelled data. Experiments show that this method has high classification accuracy, less computations and strong robustness. It is a new semi-supervised method for classification of PQDs. Show more

Keywords: Deep learning, generative adversarial network, power quality, semi-supervised learning, signal classification.

DOI: 10.3233/JIFS-191274

Citation: Journal of Intelligent & Fuzzy Systems, vol. 40, no. 3, pp. 3875-3885, 2021

Authors: Zhang, Ya | Xiong, Qiang

Article Type: Research Article

Abstract: The traditional method of Guangdong embroidery image color perception recognition has poor stereoscopic color reduction. Therefore, this paper introduces discrete mathematical model to design a new method of Guangdong embroidery image color perception recognition. Through histogram equalization, the input image with relatively concentrated gray distribution is transformed into the histogram output image with approximately uniform distribution to enhance the dynamic range of pixel gray value. The image of Yuexiu is smoothed and filtered by median filtering method to remove the noise in the image of Yuexiu. The RGB spatial model and HSI spatial model of image color are constructed by …normalizing the coordinates and color attributes of pixels. The RGB color space and HSI color space are transformed, and the image color perception recognition model is established to realize the color perception recognition of Guangdong embroidery image. The experimental results show that the pixels of each color in the color pixel image curve of the proposed method are as high as 800, the color pixel image curve distribution is the most intensive, and the color restoration is high. Show more

Keywords: Discrete mathematical model, Guangdong embroidery image, color perception and recognition, histogram coefficient, probability density function

DOI: 10.3233/JIFS-191484

Citation: Journal of Intelligent & Fuzzy Systems, vol. 40, no. 3, pp. 3887-3897, 2021

Authors: Chen, Yanhao | Yu, Suihuai | Chu, Jianjie | Yu, Mingjiu | Chen, Dengkai

Article Type: Research Article

Abstract: Numerous human factors need to be considered in the analysis and design of aircraft cockpits. The investigation of cockpit color patterns and the design of matched color schemes are particularly important. In this paper, we propose a fuzzy-logic-based emotional evaluation scheme for color pattern design in aircraft cockpits. Color pattern samples were collected and analyzed to construct a color library for cockpit color matching. Color scheme groups were accordingly redesigned and visualized. Based on fuzzy-set theory, the Kansei engineering model was thus used to evaluate the emotional image of the color schemes, and rank them in terms of priority. A …support vector machine was trained to construct a comprehensive intelligent color evaluation system. After training and validating the evaluation model, accurate emotional evaluation of color matching schemes could be achieved. Thus, the proposed system enables the extraction, mapping and evaluation of cockpit color matching schemes, and can be used in color scheme design for other cabins. Show more

Keywords: Emotional evaluation, aircraft cockpit design, color matching, fuzzy TOPSIS, intelligent design

DOI: 10.3233/JIFS-191960

Citation: Journal of Intelligent & Fuzzy Systems, vol. 40, no. 3, pp. 3899-3917, 2021

Authors: Hei, Hongzhong | Jian, Xianzhong | Xiao, Erliang

Article Type: Research Article

Abstract: The widespread application of infrared human action recognition in intelligent surveillance has attracted significant attention. However, the infrared action recognition dataset is limited, which limits the development of infrared action recognition. Existing methods for infrared action recognition are based on features in the same sample, without paying attention to within-class differences. Motivated by the idea of weighting video information, this paper proposes a novel infrared action recognition framework to reweight the samples of training sets named REWS to solve the problems of limited infrared action data and the large within-class differences in the infrared action recognition dataset. In the proposed …framework, we first map infrared action video data to a low-dimensional feature space, and use the cosine similarity between the feature data of the training set and the testing set to determine the weight of the training set samples. Each training set sample has an independent weight. Then, a support vector machine (SVM) is trained by the training sets with weights to recognize the infrared actions. Experimental results demonstrate that our approach can achieve state-of-the-art performance compared with hand-crafted features based methods on the benchmark InfAR dataset. Show more

Keywords: Infrared, action recognition, within-class differences, samples reweighting, cosine similarity

DOI: 10.3233/JIFS-192068

Citation: Journal of Intelligent & Fuzzy Systems, vol. 40, no. 3, pp. 3919-3930, 2021

Authors: Li, Yundong | Liu, Yi | Dong, Han | Hu, Wei | Lin, Chen

Article Type: Research Article

Abstract: The intrusion detection of railway clearance is crucial for avoiding railway accidents caused by the invasion of abnormal objects, such as pedestrians, falling rocks, and animals. However, detecting intrusions using deep learning methods from infrared images captured at night remains a challenging task because of the lack of sufficient training samples. To address this issue, a transfer strategy that migrates daytime RGB images to the nighttime style of infrared images is proposed in this study. The proposed method consists of two stages. In the first stage, a data generation model is trained on the basis of generative adversarial networks using …RGB images and a small number of infrared images, and then, synthetic samples are generated using a well-trained model. In the second stage, a single shot multibox detector (SSD) model is trained using synthetic data and utilized to detect abnormal objects from infrared images at nighttime. To validate the effectiveness of the proposed method, two groups of experiments, namely, railway and non-railway scenes, are conducted. Experimental results demonstrate the effectiveness of the proposed method, and an improvement of 17.8% is achieved for object detection at nighttime. Show more

Keywords: Railway clearance, infrared image detection, CycleGAN, SSD

DOI: 10.3233/JIFS-192141

Citation: Journal of Intelligent & Fuzzy Systems, vol. 40, no. 3, pp. 3931-3943, 2021

Authors: Liao, Yu-Hsien

Article Type: Research Article

Abstract: In real situations, players might represent administrative areas of different scales; players might have different activity abilities. Thus, we propose an extension of the Banzhaf-Owen index in the framework of fuzzy transferable-utility games by considering supreme-utilities and weights simultaneously, which we name the weighted fuzzy Banzhaf-Owen index. Here we adopt three existing notions from traditional game theory and reinterpret them in the framework of fuzzy transferable-utility games. The first one is that this weighted index could be represented as an alternative formulation in terms of excess functions. The second is that, based on an reduced game and related consistency, we …offer an axiomatic result to present the rationality of this weighted index. Finally, we introduce two dynamic processes to illustrate that this weighted index could be reached by players who start from an arbitrary efficient payoff vector and make successive adjustments. Show more

Keywords: Weight, the weighted fuzzy Banzhaf-Owen index, excess function, dynamic process, 91A, 91B

DOI: 10.3233/JIFS-192165

Citation: Journal of Intelligent & Fuzzy Systems, vol. 40, no. 3, pp. 3945-3956, 2021