An Efficient Defect Classification Algorithm for Ceramic Tiles

Ragab, Khaled; Alsharay, Nahed

doi:10.1007/978-3-319-56660-3_21

Cited by 4 publications

(6 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Ragab and Alsharay 6 detailed various kinds of tile surface defects including crack, bumps, depressions, holes, dirt, drop, water drops, and so forth. They also proposed an algorithm to detect spots and cracks as follows: (1) preprocess the test image, and then (2) detect defects in block, as well as (3) classify the types of the defects.…”

Section: Related Workmentioning

confidence: 99%

“…Moreover, the positive and negative samples are trained alternately and data set is shuffled in each iteration. The cross-entropy loss function is defined as Equation (6).…”

Section: Training Detailsmentioning

confidence: 99%

“…To address this issue, a number of researchers proposed various methods. [2][3][4][5][6][7][8][9][10][11][12][13] The common ideas of these methods are that filtering operators, wavelet transform, or morphological operations are first facilitated to extract hand-crafted features from captured images, and then classifiers (e.g., support-vector machine and Decision Tree) are employed to classify the defects. The accuracy of these methods directly depends on the hand-crafted extracted features.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

HFENet: A lightweight hand‐crafted feature enhanced CNN for ceramic tile surface defect detection

Zhang

Guo

et al. 2022

Int J of Intelligent Sys

View full text Add to dashboard Cite

Inkjet printing technology can make tiles with very rich and realistic patterns, so it is widely adopted in the ceramic industry. However, the frequent nozzle blockage and inconsistent inkjet volume by inkjet printing devices, usually leads to defects such as stayguy and color blocks in the tile surface. Especially, the stayguy in complex pattern is difficult to identify by naked eyes due to it is easily covered by complex patterns and becomes invisible, this brings great challenge to tile quality inspection. Nowadays, the machine learning is employed to address the issues. The existing machine learning methods based on hand‐crafted features are capable of stayguy detection of the tiles with a simple pattern, but not applicable for complex patterns due to the interference of pattern in feature extraction. The emerging deep‐learning‐based methods have the potential to be applied for stayguy detection with complex patterns, but cannot achieve real‐time detection due to high complexity. In this paper, a lightweight hand‐crafted feature enhanced convolutional neural network (named HFENet) is proposed for rapid defect detection of tile surface. First, we perform data enhancement on the original image by global histogram equalization and image addition. Second, for the special shape of stayguy which is usually vertical, we embed the extended vertical edge detection operator (Prewitt) as convolution kernel into HFENet to extract the hand‐crafted vertical edge features of the test image and eliminate the interference of complex pattern in the feature extraction. Third, the 5 × 1 asymmetric convolution kernel with a dilation rate of 2 is used to improve the utilization of convolution kernel and reduce the complexity of the model. Fourth, to reach the real‐time requirements, a memory access cost‐aware design is proposed, which can orchestrate the number of shallow convolution layers and deep convolution layers in feature extraction. The experiments were performed on the ceramic tile image data set captured by high‐resolution industrial cameras in ceramic tile production line. Experimental results show that the HFENet outperforms the state‐of‐the‐art semantic segmentation networks (i.e., UNet, FCN‐8s, SegNet, DeepLabV3+, etc.) and lightweight networks (i.e., ShuffleNet, MobileNet, and SqueezeNet). All the code and data are available at a GitHub repository (https://github.com/RobotvisionLab/HFENet).

show abstract

Section: Related Workmentioning

confidence: 99%

“…Moreover, the positive and negative samples are trained alternately and data set is shuffled in each iteration. The cross-entropy loss function is defined as Equation (6).…”

Section: Training Detailsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

HFENet: A lightweight hand‐crafted feature enhanced CNN for ceramic tile surface defect detection

Zhang

Guo

et al. 2022

Int J of Intelligent Sys

View full text Add to dashboard Cite

show abstract

“…Penelitian yang dilakukan [5] mengusulkan suatu algoritma yang membagi gambar keramik ke dalam partisi dan mengidentifikasi partisi yang cacat. Selanjutnya, algoritma klasifikasi hanya diterapkan pada partisi yang cacat.…”

Section: Pendahuluanunclassified

“…, Correlation (Cor) yang dinyatakan dengan persamaan (4), Homogeneity atau Inverse Difference Moment (IDM) yang dinyatakan dengan persamaan(5).…”

unclassified

Deteksi Cacat Ubin Keramik Dengan Metode K-Nearest Neighbor

Alamsyah

Wiranata

Rafie

2019

View full text Add to dashboard Cite

AbstrakPerusahaan industri manufaktur harus dapat menjaga kualitas dari setiap produk yang diproduksi, termasuk perusahaan industri manufaktur yang memproduksi ubin keramik. Selama beberapa tahun, inspeksi visual secara otomatis sudah diterapkan untuk menentukan kualitas ubin keramik yang diproduksi. Sulitnya mendeteksi ubin keramik yang cacat bisa berdampak pada menurunnya kualitas hasil produksi, menurunnya tingkat kepercayaan konsumen, dan penurunan laba bagi perusahaan. Masalah yang dibahas di dalam penelitian ini adalah bagaimana mendeteksi ubin keramik yang cacat sehingga model yang dibangun dapat meningkatkan akurasi untuk mendeteksi ubin keramik yang cacat. Langkah penyelesaian masalah ini adalah dengan mengumpulkan data berupa citra dari ubin keramik, kemudian data citra dilakukan preprocessing menggunakan Median Filtering untuk menghilangkan noise salt and paper dan Teknik Morfologi untuk memperbaiki hasil segmentasi citra. Setelah dilakukan preprocessing, data citra diekstraksi ciri berdasarkan tekstur dengan menggunakan metode Gray Level Co-occurrence Matrix (GLCM) yang dilanjutkan dengan mengklasifikasikan data citra menggunakan metode K-Nearest Neighbor (KNN). Hasil dari penelitian ini adalah model yang dibangun menggunakan metode K-Nearest Neighbor dapat meningkatkan akurasi untuk mendeteksi kecacatan pada ubin keramik dengan nilai akurasi sebesar 98.9474% untuk k = 3. Kata kunci-Digital Image Processing¸ Median Filtering, Teknik Morfologi, GLCM dan KNN AbstractManufacturing industry companies must be able to maintain the quality of each product produced, including manufacturing companies that produce ceramic tiles. For several years, automatic visual inspection has been applied to determine the quality of ceramic tiles produced. The difficulty of detecting defective ceramic tiles can have an impact on decreasing the quality of production, decreasing the level of consumer confidence, and decreasing profits for the company. The problem discussed in this research is how to defect detection of ceramic tiles so that the model built can improve accuracy to defect detection of ceramic tiles. The solution to this problem is to collect data in the form of ceramic tiles images, then preprocessing images data using Median Filtering to eliminate salt and paper noise and Morphological Techniques to improve images segmentation results. After preprocessing, texture image extraction data is based on texture using the Gray Level Co-occurrence Matrix (GLCM) method which is continued by classifying images data using the K-Nearest Neighbor (KNN) method. The results of this research are models that are built using the K-Nearest Neighbor method can improve accuracy to defect detection of ceramic tiles with an accuracy value of 98.9474% for k=3.

show abstract