Fabric image retrieval algorithm based on fractal coding and Zernike moment under the wavelet transform
doi:10.37188/CO.EN-2022-0021
-
摘要:
为帮助纺织企业的工作人员快速、准确地从数据库中检索出与织物图像相同或相似的图像,提出了一种小波变换下基于分形编码和 Zernike 矩的织物图像检索算法。首先,利用小波变换获得低频分量,对变换后的低频子图进行分形编码,得到编码参数。然后,计算低频子图像的 Zernike 矩。将小波变换下的分形编码参数和Zernike 矩相结合作为织物图像检索的特征量。相比于单特征检索方法,该算法克服了精度低、耗时长的问题。与基本分形算法(BFIC)、联合正交分形参数和改进的 Hu 不变矩算法(HVKF)以及稀疏分形图像压缩算法(SFIC)相比,该算法确保了重建图像的质量和较低的编码时间。实验结果表明,织物图像检索的平均精度和平均召回率均高于现有的检索方法。
Abstract:A fabric image retrieval algorithm based on fractal coding and Zernike moments under a wavelet transform is proposed, which can quickly and accurately retrieve images from a database that are similar to fabric images submitted for retrieval. Firstly, the low-frequency component is obtained by a wavelet transform, and the transformed low-frequency sub-image is fractally encoded to obtain its coding parameters. Then, the Zernike moment of the low-frequency sub-image is calculated. The fractal coding parameters and Zernike moment under a wavelet transform are combined as the fabric image retrieval characteristic. The algorithm overcomes the problems of low retrieval accuracy and the high time consumption of direct feature extraction under a single feature. Compared with the Basic Fractal Image Compression (BFIC) algorithm, the joint orthogonal fractal parameters with the improved Hu invariant moment and Variable bandwidth Kernel density estimation of Fractal parameters (HVKF) algorithm and the Sparse Fractal Image Compression (SFIC) algorithm, the proposed algorithm ensures the quality and lower encoding time of the reconstructed image. The experiments show that the average precision and average recall of fabric image retrieval are higher than those of existing methods.
-
Key words:
- fabric image retrieval/
- wavelet transform/
- fractal coding/
- Zernike moment
-
Table 1.Isometric transform
j $q(j)$ 1 Identity transformation 2 symmetry of theXaxis 3 symmetry of theYaxis 4 Rotate 180 degrees 5 $y = - x$ 6 $y = x$ 7 Rotate 90 degrees counterclockwise 8 Rotate 270 degrees counterclockwise Table 2.Average PSNR of 3000 images with four different methods
Method BFIC HVKF SFIC FZW PSNR/dB 28.26 31.47 36.38 37.21 Table 3.Comparison of decoding image quality and encoding time under different algorithms
Images BFIC HVKF SFIC FZW PSNR/dB Time/s SSIM PSNR/dB Time/s SSIM PSNR/dB Time/s SSIM PSNR/dB Time/s SSIM Trellis 28.44 727.81 0.805 32.72 165.37 0.852 37.86 65.76 0.938 35.89 43.67 0.921 Flower1 27.72 748.32 0.742 31.85 148.88 0.823 38.53 83.63 0.955 38.82 43.85 0.978 Cluster 29.01 733.70 0.846 30.46 160.53 0.869 35.48 90.08 0.937 36.30 38.23 0.945 Stripes1 28.56 742.24 0.784 30.80 156.60 0.858 36.06 82.34 0.943 36.71 38.13 0.960 Leaves 28.99 736.59 0.808 33.54 163.47 0.889 37.25 71.09 0.946 37.57 42.97 0.966 Stripes2 29.12 740.68 0.812 29.23 163.26 0.856 37.93 87.72 0.933 38.15 44.28 0.974 Flower2 28.70 728.76 0.774 30.53 155.04 0.842 38.22 73.51 0.986 37.64 38.11 0.982 Rhombus 27.85 730.05 0.692 30.07 163.77 0.870 33.10 82.55 0.969 35.21 42.93 0.983 Flame 28.13 757.90 0.802 31.07 169.90 0.858 36.37 80.61 0.944 37.29 47.57 0.975 Diamond 27.32 724.81 0.769 30.32 166.23 0.810 34.75 78.85 0.926 38.43 51.52 0.979 Curve 28.66 675.55 0.807 32.45 147.96 0.871 36.18 63.48 0.932 37.06 35.92 0.968 Dots 29.57 701.53 0.821 32.76 158.03 0.864 36.92 78.80 0.939 38.28 37.69 0.986 Wave 28.29 681.63 0.794 30.97 150.54 0.806 37.15 76.05 0.945 37.18 38.50 0.964 Scroll 27.73 717.84 0.654 29.60 161.86 0.797 35.10 79.37 0.899 36.99 38.03 0.943 Twill1 29.54 727.54 0.811 31.55 163.43 0.853 37.22 75.91 0.938 37.61 39.44 0.955 Circle1 27.94 720.09 0.763 31.89 159.87 0.847 34.88 73.00 0.917 37.26 40.16 0.967 -
[1] SINGH C, KAUR K P. A fast and efficient image retrieval system based on color and texture features[J].Journal of Visual Communication and Image Representation, 2016, 41: 225-238.doi:10.1016/j.jvcir.2016.10.002 [2] LEE J, SUL I. Construction of garment pattern shape information system using image analysis and shape recognition techniques[J].International Journal of Clothing Science and Technology, 2016, 28(4): 543-555.doi:10.1108/IJCST-10-2015-0114 [3] HU X D, FU M Y, ZHU ZH J,et al. Unsupervised defect detection algorithm for printed fabrics using content-based image retrieval techniques[J].Textile Research Journal, 2021, 91(21-22): 2551-2566.doi:10.1177/00405175211008614 [4] ZHANG N, XIANG J, WANG L,et al. Image retrieval of wool fabric. Part II: based on low-level color features[J].Textile Research Journal, 2020, 90(7-8): 797-808.doi:10.1177/0040517519881819 [5] XIANG J, ZHANG N, PAN R R,et al. Fabric retrieval based on multi-task learning[J].IEEE Transactions on Image Processing, 2021, 30: 1570-1582.doi:10.1109/TIP.2020.3043877 [6] JIANG D Y, KIM J. Image retrieval method based on image feature fusion and discrete cosine transform[J].Applied Sciences, 2021, 11(12): 5701.doi:10.3390/app11125701 [7] KHALID M J, IRFAN M, ALI T,et al. Integration of discrete wavelet transform, DBSCAN, and classifiers for efficient content based image retrieval[J].Electronics, 2020, 9(11): 1886.doi:10.3390/electronics9111886 [8] PAN R R, GAO W D, LI W,et al. Image analysis for seam-puckering evaluation[J].Textile Research Journal, 2017, 87(20): 2513-2523.doi:10.1177/0040517516673330 [9] LIU P ZH, GUO J M, CHAMNONGTHAI K,et al. Fusion of color histogram and LBP-based features for texture image retrieval and classification[J].Information Sciences, 2017, 390: 95-111.doi:10.1016/j.ins.2017.01.025 [10] XIN S, SONG ZH G, SHI J L,et al. Multiple channels local binary pattern for color texture Representation and classification[J].Signal Processing:Image Communication, 2021, 98: 116392.doi:10.1016/j.image.2021.116392 [11] JAMIL N, SOH H C, SEMBOK T M T,et al.. A modified edge-based region growing segmentation of geometric objects[C]//Lecture Notes in Computer Science. Berlin: Springer-Verlag, 2011: 99. [12] FU B L, LIU X G. An intelligent computational framework for the definition and identification of the womenswear silhouettes[J].International Journal of Clothing Science and Technology, 2019, 31(2): 158-180.doi:10.1108/IJCST-08-2017-0128 [13] CORPUS G, PIÑERO D P. Short-term effect of wearing of extended depth-of-focus contact lenses in myopic children: a pilot study[J].Applied Sciences, 2022, 12(1): 431.doi:10.3390/app12010431 [14] BAR O, BIBRZYCKI Ł, NIEDŹWIECKI M,et al. Zernike moment based classification of cosmic ray candidate hits from CMOS sensors[J].Sensors, 2021, 21(22): 7718.doi:10.3390/s21227718 [15] YU X L, WANG H L. Support vector machine classification model for color fastness to ironing of vat dyes[J].Textile Research Journal, 2021, 91(15-16): 1889-1899.doi:10.1177/0040517521992366 [16] FAYAZ M, TOROKELDIEV N, TURDUMAMATOV S,et al. An efficient methodology for brain MRI classification based on DWT and convolutional neural network[J].Sensors, 2021, 21(22): 7480.doi:10.3390/s21227480 [17] DARAEE F, MOZAFFARI S. Watermarking in binary document images using fractal codes[J].Pattern Recognition Letters, 2014, 35: 120-129.doi:10.1016/j.patrec.2013.04.022 [18] AHMAD M, AGARWAL S, ALKHAYYAT A,et al. An image encryption algorithm based on new generalized fusion fractal structure[J].Information Sciences, 2022, 592: 1-20.doi:10.1016/j.ins.2022.01.042 [19] JAGANNADHAM D B V, RAJU G V S, NARAYANA D V S. Novel performance analysis of DCT, DWT and fractal coding in image compression[M]//RAJU K S, SENKERIK R, LANKA S P,et al.. Data Engineering and Communication Technology. Singapore: Springer, 2020: 611-622. [20] HUANG X Q, ZHANG Q, LIU W B. A new method for image retrieval based on analyzing fractal coding characters[J].Journal of Visual Communication and Image Representation, 2013, 24(1): 42-47.doi:10.1016/j.jvcir.2012.10.005 [21] ZHANG Q, HUANG X Q, LIU W B,et al. An effective image retrieval method based on Kernel Density Estimation of collage error and moment invariants[J].Journal of Electronics(China), 2013, 30(4): 391-400.doi:10.1007/s11767-013-3031-4 [22] TEAGUE M R. Image analysis via the general theory of moments[J].Journal of the Optical Society of America, 1980, 70(8): 920-930. [23] WANG Y, ZHAO Y SH, CHEN Y. Texture classification using rotation invariant models on integrated local binary pattern and Zernike moments[J].Eurasip Journal on Advances in Signal Processing, 2014, 2014(1): 182.doi:10.1186/1687-6180-2014-182 [24] SWAIN M, SWAIN D. An effective watermarking technique using BTC and SVD for image authentication and quality recovery[J].Integration, 2022, 83: 12-23. [25] WANG ZH, BOVIK A C, SHEIKH H R,et al. Image quality assessment: from error visibility to structural similarity[J].IEEE Transactions on Image Processing, 2004, 13(4): 600-612.doi:10.1109/TIP.2003.819861 [26] WU ZH H, SONG T T, ZHANG Y B. Quantumk-means algorithm based on Manhattan distance[J].Quantum Information Processing, 2022, 21(1): 19.doi:10.1007/s11128-021-03384-7 [27] JIANG X P, HU X H, HE T T. Identification of the clustering structure in microbiome data by density clustering on the Manhattan distance[J].Science China Information Sciences, 2016, 59(7): 070104.doi:10.1007/s11432-016-5587-8 [28] FU G H, XU F, ZHANG B Y,et al. Stable variable selection of class-imbalanced data with precision-recall criterion[J].Chemometrics and Intelligent Laboratory Systems, 2017, 171: 241-250.doi:10.1016/j.chemolab.2017.10.015 [29] ZHANG Q, LIN Q H, KANG X. Research on image retrieval based on kernel density estimation and fractal coding algorithm[J].Acta Metrologica Sinica, 2017, 38(3): 284-287. (in Chinese)doi:10.3969/j.issn.1000-1158.2017.03.07 [30] WANG J J, CHEN P, XI B,et al. Fast sparse fractal image compression[J].PLoS One, 2017, 12(9): e0184408.doi:10.1371/journal.pone.0184408 [31] ZHA T. Application comparison of textile fabric image retrieval algorithms based on content[J].Journal of Textile Science&Fashion Technology, 2020, 7(2): 659.