Fuzzy entropy and image thresholding are the most direct and effective methods for image segmentation. This paper, taking fuzzy Kapur’s entropy as the optimal objective function, with modified discrete Grey wolf optimizer (GWO) as the tool, uses pseudotrapezoid-shaped to conduct fuzzy membership initialization so as to achieve image segmentation finally by means of local information aggregation. Experiment results show that the proposed fuzzy-based GWO and aggregation algorithm and fuzzy-based modified discrete GWO and aggregation (FMDGWOA) algorithm can search out the optimal thresholds effectively and accurately. In this paper, electro-magnetism optimization based on Kapur’s entropy, standard GWO and fuzzy entropy-based differential evolution algorithm are experimentally compared with the proposed method, respectively. It shows that FMDGWOA enjoys obvious advantages in segmentation quality, objective function, and stability.

Fuzzy Multilevel Image Thresholding Based on Modified Discrete Grey Wolf Optimizer and Local Information Aggregation

Some image’s regions have unbalance information, such as blurred contour, shade, and uneven brightness. Those regions are called as ambiguous regions. Ambiguous region cause problem during region merging process in interactive image segmentation because that region has double information, both as object and background. We proposed a new region merging strategy using fuzzy similarity measurement for image segmentation. The proposed method has four steps; the first step is initial segmentation using mean-shift algorithm. The second step is giving markers manually to indicate the object and background region. The third step is determining the fuzzy region or ambiguous region in the images. The last step is fuzzy region merging using fuzzy similarity measurement. The experimental results demonstrated that the proposed method is able to segment natural images and dental panoramic images successfully with the average value of misclassification error (ME) 1.96% and 5.47%, respectively.

/pdf/fuzzy-region-merging-using-fuzzy-similarity-measurement-on-1ljzdul1z5.pdf

Fuzzy Region Merging using Fuzzy Similarity Measurement on Image Segmentation

The non-contact measuring and segmentation method is of primary concern for heavy forging industrial applications. Such a measuring method, if successful, can improve the productivity and quality of manufactured items, reduce human labor, minimize material waste, and even further permit real-time control. This proposed method concerns with the utilization a fuzzy logic automatic analysis thermal images as of industrial products for defect detection. An original two stages algorithm is presented based on the feature analysis of the thermal images obtained from the inspected product. Each object in the image is analyzed using fuzzy logic techniques. The first stage takes an automatic decision whether the current object can be classified as a defect from the geometrical point of view and the second stage takes the final decision by using “logical” criteria that are dependent on the product at hand and its quality requirement, These results indicate that the proposed approach is more practical and convenient than current methods for real-time, on-site measurements.

/pdf/a-multi-level-fuzzy-threshold-image-segmentation-method-for-2cd1ut63vg.pdf

A Multi Level Fuzzy Threshold Image Segmentation Method for Industrial Applications

Dalam segmentasi citra, thresholding merupakan salah metode yang mudah dan sederhana untuk diimplementasikan. Pada citra panoramik gigi, penentuan global threshold masih kurang begitu optimal untuk diimplementasikan. Hal tersebut dikarenakan adanya factor penghambat seperti pencahayaan yang tidak merata dan citra yang kabur. Faktor-faktor tersebutn dapat menyebabkan histogram tidak bisa dipartisi dengan baik, sehingga akan berpengaruh pada hasil segmentasi. Pada penelitian ini diusulkan lokal fuzzy thresholding berdasarkan pengukuran fuzzy similarity pada interaktif segmentasi citra panoramik gigi. Metode yang diusulkan terdiri dari tiga tahapan utama, tahap pertama region splitting untuk mendapatkan lokal region. Tahap kedua adalah user marking untuk mendapat inisial seed background dan objek, Tahap terakhir adalah pengukuran fuzzy similarity pada setiap lokal region untuk mendapatkan nilai lokal threshold. Hasil uji coba pada citra panoramik gigi, metode yang diusulkan berhasil melakukan segmentasi dengan rata-rata missclasification error (ME) 5.47%.

/pdf/lokal-fuzzy-thresholding-berdasarkan-pengukuran-fuzzy-la1d2cpip8.pdf

Lokal Fuzzy Thresholding Berdasarkan Pengukuran Fuzzy Similarity Pada Interaktif Segmentasi Citra Panoramik Gigi

Binary Segmentation of an image played an important role in many image processing application. An image that was having no bimodal (or nearly) histogram accompanied by low-contrast was still a challenging segmentation problem to address. In this paper, we proposed a new segmentation strategy to images with very irregular histogram and had not significant contrast using index of fuzziness and adaptive thresholding. Index of fuzziness was used to determine the initial threshold, while adaptive thresholding was used to refine the coarse segmentation results. The used data were grayscale images from related papers previously. Moreover, the proposed method would be tested on the grayscale images of malaria parasite candidates from thickblood smear that had the same problem with this research. The experimental results showed that the proposed method achieved higher segmentation accuracy and lower estimation error than other methods. The method also effective proven to segment malaria parasite candidates from thickblood smears image.

/pdf/incorporating-index-of-fuzziness-and-adaptive-thresholding-2iomfn9apm.pdf

Incorporating Index of Fuzziness and Adaptive Thresholding for Image Segmentation

In this paper, we propose an automatic image thresholding method based on an index of fuzziness and a fuzzy similarity measure. This work aims at overcoming the limitation of the existing method which is semi-supervised. Using an index of fuzziness, two initial regions of gray levels located at the boundaries of the histogram are defined based on the fuzzy region. Then the threshold point is found by using a fuzzy similarity measure. No prior knowledge of the image is required. Experiments on practical images illustrate the effectiveness of the proposed method.

Zhencheng Hu

Papers

Image thresholding based on index of fuzziness and fuzzy similarity measure