Illumination Chromaticity Estimation Using Bayesian Kernel Methods

Textile Research Journal

Zhang

et al. 2018

This study proposes an ensemble differential evolution online sequential extreme learning machine (DE-OSELM) for textile image illumination correction based on the rotation forest framework. The DE-OSELM solves the inaccuracy and long training time problems associated with traditional illumination correction algorithms. First, the Grey–Edge framework is used to extract the low-dimensional and efficient image features as online sequential extreme learning machine (OSELM) input vectors to improve the training and learning speed of the OSELM. Since the input weight and hidden-layer bias of OSELMs are randomly obtained, the OSELM algorithm has poor prediction accuracy and low robustness. To overcome this shortcoming, a differential evolution algorithm that has the advantages of good global search ability and robustness is used to optimize the input weight and hidden-layer bias of the DE-OSELM. To further improve the generalization ability and robustness of the illumination correction model, the rotation forest algorithm is used as the ensemble framework, and the DE-OSELM is used as the base learner to replace the regression tree algorithm in the original rotation forest algorithm. Then, the obtained multiple different DE-OSELM learners are aggregated to establish the prediction model. The experimental results show that compared with the textile color correction algorithm based on the support vector regression and extreme learning machine algorithms, the ensemble illumination correction method achieves high prediction accuracy, strong robustness, and good generalization ability.

Section: Literature Reviewmentioning

confidence: 99%

A novel hybrid model using the rotation forest-based differential evolution online sequential extreme learning machine for illumination correction of dyed fabrics

Textile Research Journal

Zhang

et al. 2018

“…Referring to the high‐order structure of the image, Wang et al 7 proposed an illumination estimation algorithm based on support vector regression (SVR) and gray edge. Zhao et al 8 used two Bayesian kernel methods, Gaussian process regression (GPR), and correlation vector machine (RVM), to estimate the illumination chromaticity and predict the reliability of the estimation process. Martin 9 redesigned the convolution framework so that each image specific filter obtained by principal component analysis (PCA) generates a light source estimation, and then uses SVR to combine these individual estimates into a joint light source estimation.…”

Section: Introductionmentioning

confidence: 99%

Illumination estimation via random vector functional link based on improved arithmetic optimization algorithm

Liu

Color Research & Application

2021

In order to solve the problem that the existing illumination estimation methods have low accuracy in image restoration, a random vector functional link illumination estimation algorithm based on whale optimization arithmetic algorithm is proposed in this article. First, in order to improve the optimization accuracy of arithmetic optimization algorithm, the whale optimization algorithm is applied to optimize the initial population of the arithmetic optimization algorithm. Then, we utilize the improved arithmetic optimization algorithm to improve the input weight and hidden layer offset of random vector functional link, and exclude the uncertainty caused by the random generation of random vector functional link's input weight and hidden layer bias. Finally, the proposed whale arithmetic optimization algorithm‐random vector functional link model is applied to SFU lab dataset and Gehler‐Shi dataset for illumination estimation. Compared with the other earlier illumination estimation algorithms, the proposed illumination estimation model in this article has better effect and higher stability. In addition, due to the diagonal mapping method for color correction and image restoration, the proposed illumination estimation algorithm can also get better results in image restoration.

“…Supervised light estimation algorithms usually rely on traditional machine learning method, and use prior information to estimate the light of scene indirectly through the statistics and learning of light colors in a large number of known images. Zhao et al 7 proposed a Bayesian kernel method for illumination chromaticity estimation, which uses two Bayesian methods to estimate illumination chromaticity and estimate the reliability of the process. Gardei et al 8 introduced back‐propagation (BP) neural network into color constancy calculation, taking a large number of known light source images as training objects, the input image chromaticity as the input of the neural network, and the illumination chromaticity as the output to train the mapping relationship between them.…”

Section: Introductionmentioning

confidence: 99%

Dyed fabric illumination estimation with regularized random vector function link network

Color Research & Application

Dexin

Guo

et al. 2020

The unstable light source in the printing and dyeing environment would cause the change of fabric surface color and lead to serious color difference evaluation error. To solve this problem, a dyed fabric illumination estimation with regularized random vector functional link network (RRVFL) was proposed in this study. First, the Gray‐Edge framework is used to extract the color features of the sample images collected from the actual scene, and the extracted color features and the illumination information of the sample images constitute a data set. Then, considering the ill‐conditioned solution of the output weight of the traditional random vector functional link network (RVFL), regularization is proposed to solve this problem. Thus, we constitute a highly robust RRVFL dyed fabric illumination model. By analyzing the parameters that affect the precision of RRVFL model, the optimal parameters of RRVFL can be selected. Finally, the traditional RVFL, extreme learning machine (ELM), back‐propagation (BP), regularized extreme learning machine (RELM), support vector regression (SVR), and RRVFL estimation algorithm proposed in this study were compared and analyzed using the measurement standards of angle error, colorimetric error, and T‐test, respectively. The experimental results show that RRVFL has the best predictive results and the most stable performance compared with the traditional algorithm. RRVFL is 0.00036, 2.8050, 3.3518, 4.1669, and 2.9289 less than RVFL, ELM, BP, RELM, and SVR, respectively, in terms of average angle error. RRVFL is 0.0131, 0.0763, 0.0232, 0.0241, and 0.0221 lower than RVFL, ELM, BP, RELM, and SVR, respectively, in the average colorimetric error.