Estimation and monitoring of product aesthetics: application to manufacturing of “engineered stone” countertops

Computers & Chemical Engineering

2014

The monitoring, fault detection and visualization of defects are a strategic issue for product quality. This paper presents a novel methodology based on the integration of textural Multivariate image analysis (MIA) and multivariate statistical process control (MSPC) for process monitoring. The proposed approach combines MIA and p-control charts, as well as T 2 and RSS images for defect location and visualization. Simulated images of steel plates are used to illustrate the monitoring performance of it. Both approaches are also applied on real clover images. Original Research Article2

Section: Pfpm Approachmentioning

confidence: 97%

Section: Competing Approachmentioning

confidence: 99%

Statistical process control based on Multivariate Image Analysis: A new proposal for monitoring and defect detection

Computers & Chemical Engineering

2014

“…This strategy lets to perform not only a classification, but also a defect detection task by using just one technique, since we are maintaining the pixel domain, instead of having to apply some other techniques from traditional imaging [12]. It implies to unfold each image into just one column vector having the intensity levels of the pixels, and registering, for each one of them, the intensities of the neighbouring ones, by following an established order in the neighbourhood.…”

Section: Image Data Structurementioning

confidence: 98%

“…In the multivariate statistical process control (MSPC) field, different works have treated the monitoring and classification problems through multivariate image analysis (MIA) [4] by using spectral information [5][6][7][8] or spatial information [9][10][11][12]. The main characteristic of MIA is the ability to analyse each image in the pixel domain.…”

Section: Introductionmentioning

confidence: 99%

Integration of colour and textural information in multivariate image analysis: defect detection and classification issues

Journal of Chemometrics

2007

In industrial processes, the detection and visualisation of defects and the development of efficient automated classification tools are strategic issues, especially when dealing with random colour textures (RCTs). This paper discusses the benefits of integrating colour and spatial (i.e. textural) information of digital RGB colour images in multivariate image analysis (MIA) to deal with these topics. Regarding the first one, a simple and computational cost-effective monitoring procedure based on colour-textural MIA merged with multivariate statistical process control (MSPC) ideas is outlined. Two novel computed images: T 2 and RSS Images are proposed. The procedure is applied on digital RGB colour images from artificial stone plates. With respect to the second issue, when colour-textural MIA is used for image classification a lot of factors (e.g. pre-processing, modelling, . . .) likely affecting the success rate in the classification (SRC) show up. This paper presents a methodology based on the combination of experimental design and logistic regression for choosing the best combination of factors to maximise the SRC of different types of images. Digital RGB colour images from ceramic tiles and orange fruits are used to illustrate the potential of the proposed methodology.

“…At present, several uses of MIA are reported in literature for different tasks, all of which are characterized by being well described by the 2 main sources of information an image can carry. Textural variability, which can be gathered by analyzing the “2” dimension relationship structure of pixels and “spectral” property variability, which is based on the “third” dimension, is the channels acquired for each pixel.…”

Section: Introductionmentioning

confidence: 99%

Coupling 2D‐wavelet decomposition and multivariate image analysis (2D WT‐MIA)

Vigni

Journal of Chemometrics

et al. 2017

The use of 2D discrete wavelet transform in the feature enhancement phase of multivariate image analysis is discussed and implemented in a comparative way with respect to previous publications. In the proposed approach, all the resulting subimages obtained by discrete wavelet transform decomposition are unfolded pixel-wise and midlevel data fused to a feature matrix that is used for the feature analysis phase. Congruent subimages can be obtained either by reconstruction of each decomposition block to the original pixel dimensions or by using the stationary wavelet transform decomposition scheme. The main advantage is that all possible relationships among blocks, decomposition levels, and channels are assessed in a single multivariate analysis step (feature analysis). This is particularly useful in a monitoring context where the aim is to build multivariate control charts based on images. Moreover, the approach can be versatile for contexts where several images are analyzed at a time as well as in the multispectral image analysis.Both a set of simple artificial images and a set of real images, representative of the on-line quality monitoring context, will be used to highlight the details of the methodology and show how the wavelet transform allows extracting features that are informative of how strong the texture of the image is and in which direction it varies.