Dyed fabric illumination estimation with regularized random vector function link network

Zhou, Zhiyu; Dexin, Liu; Guo, Jiushen; Zhang, Jianxin; Zhu, Zefei; Wang, Chao

doi:10.1002/col.22602

Cited by 15 publications

(12 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Therefore, the training takes less time and has attracted widespread attention. In order to solve the ill-conditioned solution of the output weights of RVFL, a regularized RVFL 13 method was proposed, thereby constructing a highly robust regularized RVFL dyed fabric illumination estimation model, but the randomness of input weight and hidden layer bias has not been solved yet. Because the input weight and hidden layer bias are randomly determined, the prediction accuracy of RVFL is inevitably affected by the random input weight and hidden layer bias.…”

Section: Introductionmentioning

confidence: 99%

Color constancy computation for dyed fabrics via improved marine predators algorithm optimized random vector functional‐link network

Liu

Yang

2021

Color Research & Application

View full text Add to dashboard Cite

Aiming at the color characteristics of dyed fabrics that are easily affected by changes in light, which affects the correctness of the color difference classification of dyed fabrics, this article proposes the color constancy calculation of dyed fabrics based on improved marine predators algorithm optimized random vector functional-link. First, in order to obtain the excellent initial population of the marine predators algorithm, this article uses two update strategies of the sine and cosine algorithm to screen the randomly initialized population of the marine predators algorithm. Then, the MPA algorithm that initializes the population using the sine and cosine algorithm optimizes the input weights and hidden layer bias parameters of random vector functional-link, thereby improving the prediction accuracy of random vector functional-link. Finally, using the image features extracted by the Gray-Edge framework, the sine and cosine algorithmmarine predators algorithm-random vector functional-link model proposed in this article is used to calculate the color constancy of dyed fabrics, to eliminate the influence of illumination changes on the color difference classification of dyed fabrics. Compared with the other eight algorithms, the dyed fabric image restored by the sine and cosine -marine predators -random vector functional-link algorithm proposed in this article is closest to the image under standard illumination, that is, the color constancy evaluation effect of the dyed fabric is the best.

show abstract

Section: Introductionmentioning

confidence: 99%

Color constancy computation for dyed fabrics via improved marine predators algorithm optimized random vector functional‐link network

Liu

Yang

2021

Color Research & Application

View full text Add to dashboard Cite

show abstract

“…Hailiang et al [13] performed feature extraction on face data and fed the feature set into the RVFL model for recognition, which led to a large improvement in accuracy, sparsity and stability. Zhou [14] proposed a regularized random vector function linkage (RRVFL) chromophore light estimation algorithm, which improved the prediction accuracy of RVFL, but still did not fundamentally solve the parameter randomness of RVFL caused by the algorithm instability problem.…”

Section: Introductionmentioning

confidence: 99%

An Improved GWO Algorithm Optimized RVFL Model for Oil Layer Prediction

Xia

Pan

et al. 2021

Electronics

View full text Add to dashboard Cite

In this study, a model based on the improved grey wolf optimizer (GWO) for optimizing RVFL is proposed to enable the problem of poor accuracy of Oil layer prediction due to the randomness of the parameters present in the random vector function link (RVFL) model to be addressed. Firstly, GWO is improved based on the advantages of chaos theory and the marine predator algorithm (MPA) to overcome the problem of low convergence accuracy in the optimization process of the GWO optimization algorithm. The improved GWO algorithm was then used to optimize the input weights and implicit layer biases of the RVFL network model so that the problem of inaccurate and unstable classification of RVFL due to the randomness of the parameters was avoided. MPA-GWO was used for comparison with algorithms of the same type under a function of 15 standard tests. From the results, it was concluded that it outperformed the algorithms of its type in terms of search accuracy and search speed. At the same time, the MPA-GWO-RVFL model was applied to the field of Oil layer prediction. From the comparison tests, it is concluded that the prediction accuracy of the MPA-GWO-RVFL model is on average 2.9%, 3.04%, 2.27%, 8.74%, 1.47% and 10.41% better than that of the MPA-RVFL, GWO-RVFL, PSO-RVFL, WOA-RVFL, GWFOA-RVFL and RVFL algorithms, respectively, and its practical applications are significant.

show abstract

“…Meanwhile, since the selection and setting of input weights and biases in the KELM algorithm will have a certain impact on the running effect of the KELM algorithm, it is necessary to optimize the combination of weights and bias parameters of the KELM neural network by introducing an optimization algorithm. In recent years, Zhou et al 12 proposed a hybrid illumination correction method, and Zhou et al 13 proposed illumination correction based on regularized random vector function link network. With the application of big data, a method of illumination correction based on deep learning 14,15…”

Section: Introductionmentioning

confidence: 99%

Illumination correction with optimized kernel extreme learning machine based on improved marine predators algorithm

Yang

Zheng

Chen

2021

Color Research & Application

View full text Add to dashboard Cite

Aiming at the problem of poor image illumination correction accuracy, a kernel extreme learning machine optimized based on the differential evolution‐improved marine predators algorithm is proposed to estimate the scene illumination information and restore the image. First, in order to solve the problem of randomization of the initial population of the marine predators algorithm, differential evolutions is used to provide a set of suitable initial populations to the algorithm. Then, an improved algorithm is used to optimize the weight and bias of the kernel extreme learning machine to solve the problem of the randomness of the weight and bias, so that the learning machine converges to the global optimal value and avoids the unstable predictions. Finally, after obtaining the predicted illumination information, the image model is corrected to the image effect under standard illumination through diagonal transformation. It can be seen from the experimental results that the best predictive value of the differential evolution marine predator algorithm learning machine is 0.0135915, and the quasi‐bias is only 0.001687. Compared with traditional illumination estimation algorithms such as random vector functional link and extreme learning machine, the differential evolution marine predator algorithm learning machine algorithm proposed in this article has better stability, higher accuracy of calculated predicted values, and better image restoration effect.

show abstract

Dyed fabric illumination estimation with regularized random vector function link network

Cited by 15 publications

References 25 publications

Color constancy computation for dyed fabrics via improved marine predators algorithm optimized random vector functional‐link network

Color constancy computation for dyed fabrics via improved marine predators algorithm optimized random vector functional‐link network

An Improved GWO Algorithm Optimized RVFL Model for Oil Layer Prediction

Illumination correction with optimized kernel extreme learning machine based on improved marine predators algorithm

Contact Info

Product

Resources

About