Objective Evaluation Index Research Articles

Underwater degraded image–sharpening method based on optimal polarization parameter reconstruction

In order to solve the problems of blurred image and low contrast caused by the absorption and scattering of underwater light by turbid water, an algorithm of underwater image optimal restoration based on global polarization parameter estimation was proposed. Firstly, the local minimum filter was used to estimate the underwater background light image, Stokes vector principle was introduced to calculate the degree of polarization image, and normalized mutual information was used to further optimize the degree of polarization information, to obtain the best reconstructed polarization parameters in the imaging region. Secondly, a method of image reconstruction based on morphology was used to estimate the infinite underwater background light intensity values automatically. Finally, the underwater simulated environment platform was built, and the underwater polarization image was acquired in real time through a single channel fast rotating polarization detector. In order to verify the effectiveness of the algorithm, the three objective evaluation indexes were adopted as quantification to evaluate indexes factors. The results show that the algorithm is better than the other underwater image restoration methods.

Performance Evaluation of Deep Learning Classification Network for Image Features

Deep learning (DL) has emerged as a powerful image processing technique that learns the features of the data and produces state-of-the-art prediction results. The decade from 2010 to 2020 is a real revival of DL, which has come to a turning point in history. In image classification, many deep learning networks have been proposed by scholars, and each of them has its own strengthness. It is very important and efficient for the researchers and the developers to know the performance of these networks, especially for the beginners, so as to give them a transplant instruction by an objective evaluation index. In this paper, we constructed different data sets from three aspects, texture, shape, and measurement scale to test the performance of nine mainstream image classification networks. Cross-contrast experiments were performed to analyze the sensitivity of factors which influence the stability of image classification networks. Experimental results shown that in the 27 image datasets generated by the three image factors, the classification performance of AlexNet, GoogleNet, VggNet, and DenseNet is better. The perfomance comparison of these networks are showed and discussed in details. Code and pretrained models are available at https://github.com/liqiang12689/image-classification-finall.

FSADFuse: A Novel Fusion Approach to Infrared and Visible Images

To address the problems of edge blur and weak detail resolution when fusing infrared and visible images with traditional methods, a novel image fusion approach based on fast super-resolution convolutional neural network and anisotropic diffusion is proposed. The proposed method first employs super-resolution reconstruction based on fast convolutional neural networks to get high-resolution infrared images. Second, the visible and high-resolution infrared images are anisotropically diffused to obtain the corresponding base layers and detail layers. Next, the fusion of the detail layer is performed through Karhunen-Loeve transform, and an adaptive saliency module is constructed to fuse the base layer. Finally, the fusion result is obtained by applying linear reconstruction on the fused detail layer and base layer. Compared with other algorithms, the experimental results show that the proposed algorithm not only has better objective evaluation indexes but also can better retain the details of the original image.

Beyond Camera Motion Blur Removing: How to Handle Outliers in Deblurring

Camera motion deblurring is an important low-level vision task for achieving better imaging quality. When a scene has outliers such as saturated pixels, the captured blurred image becomes more difficult to restore. In this paper, we propose a novel method to handle camera motion blur with outliers. We first propose an edge-aware scale-recurrent network (EASRN) to conduct deblurring. EASRN has a separate deblurring module that removes blur at multiple scales and an upsampling module that fuses different input scales. Then a salient edge detection network is proposed to supervise the training process and constraint the edges restoration. By simulating camera motion and adding various light sources, we can generate blurred images with saturation cutoff. Using the proposed data generation method, our network can learn to deal with outliers effectively. We evaluate our method on public test datasets including the GoPro dataset, Kohler's dataset and Lai's dataset. Both objective evaluation indexes and subjective visualization show that our method results in better deblurring quality than other state-of-the-art approaches.

Image Fusion Based on NSST and CSR Under Robust Principal Component Analysis

Aiming at the problems of loss of detail information and noise interference that are easy to produce in the image fusion process, a robust principal component analysis (RPCA) based on Convolutional Sparse Coding (CSR) and For image fusion of NonSubsampled Shear Wave Transform (NSST), the source image is pre-enhanced first; then the image is decomposed by RPCA to obtain low-rank images and sparse images; then NSST fusion is used respectively For low-rank images, CSR coding is used to fuse sparse images, and finally two separately fused images are synthesized to obtain the final fused image. Experimental results show that the algorithm in this paper can effectively improve the contrast and clarity of the fused image, reduce noise interference, rich scene information, clear targets, and overall objective evaluation indicators are better than existing algorithms, and the operating efficiency has also been improved

Empirical Study on the Impact of Environmental Protection Technology Investment on Urban Water Pollution Control Efficiency Based on Smart City Construction

In recent years, with the rapid development of science and technology, the construction of smart city has entered a new climax. But at the same time, due to the construction of smart city is still in the initial stage, there is no unified standard at home and abroad, so that the specific implementation process often leaves deficiencies, such as the more prominent problem of urban water pollution control efficiency. However, the research on this aspect is still blank in China. Therefore, this paper puts forward an empirical study on the impact of environmental protection technology investment based on smart city construction on urban water pollution control efficiency. This paper studies the core concepts of smart city construction and urban water pollution control. In order to improve the efficiency of water pollution control, we must establish a set of objective and scientific evaluation index system and matching efficiency calculation method. Therefore, this paper innovatively puts forward the empirical method of investment efficiency of environmental protection science and technology for water pollution control in smart city. Based on the current situation of smart city construction, this method classifies the factors that affect the efficiency of water pollution control, and optimizes the calculation method of empirical model. The optimized empirical model can be applied to smart cities with various economic structures, and is not limited by the size of the city, which can maximize the applicability of the model. In order to verify the actual effect of the empirical method, this paper takes Chongqing as the test object, and collects the data of the city in recent five years as the experimental sample. Through the analysis of the results of empirical tests, it can be seen that the test method of water pollution control efficiency in this paper has high accuracy, and can play an important role in water pollution control of smart city construction.

Realistic Speech-Driven Talking Video Generation with Personalized Pose

In this work, we propose a method to transform a speaker’s speech information into a target character’s talking video; the method could make the mouth shape synchronization, expression, and body posture more realistic in the synthesized speaker video. This is a challenging task because changes of mouth shape and posture are coupled with audio semantic information. The model training is difficult to converge, and the model effect is unstable in complex scenes. Existing speech-driven speaker methods cannot solve this problem well. The method proposed in this paper first generates the sequence of key points of the speaker’s face and body postures from the audio signal in real time and then visualizes these key points as a series of two-dimensional skeleton images. Subsequently, we generate the final real speaker video through the video generation network. We take a random sampling of audio clips, encode audio contents and temporal correlations using a more effective network structure, and optimize and iterate network outputs using differential loss and attitude perception loss, so as to obtain a smoother pose key-point sequence and better performance. In addition, by inserting a specified action frame into the synthesized human pose sequence window, action poses of the synthesized speaker are enriched, making the synthesis effect more realistic and natural. Then, the final speaker video is generated by the obtained gesture key points through the video generation network. In order to generate realistic and high-resolution pose detail videos, we insert a local attention mechanism into the key point network of the generated pose sequence and give higher attention to the local details of the characters through spatial weight masks. In order to verify the effectiveness of the proposed method, we used the objective evaluation index NME and user subjective evaluation methods, respectively. Experiment results showed that our method could vividly use audio contentsto generate corresponding speaker videos, and its lip-matching accuracy and expression postures are better than those of previous work. Compared with existing methods in the NME index and user subjective evaluation, our method showed better results.

A procedure to design road bioretention soil media based on runoff reduction and pollutant removal performance

Gradation design of urban road bioretention soil media is critical for the construction and performance of sponge cities, but current design methods are simplistic with reliance on local experiences. This study proposes an experiment-modeling based design procedure that effectively balances runoff reduction and pollutant removal performance based on both material properties analysis and project site performance simulation. A coarse aggregate void filling (CAVF) method is adopted for soil medium design and a stormwater treatment performance evaluation system is proposed for road bioretention containing the designed soil media. The proposed procedure was illustrated in a case study of a real project scenario. A hydrological model was built in the Storm Water Management Model (SWMM) software to evaluate the runoff reduction and pollutant removal performance. Results show that soil medium permeability is a key factor affecting the quantity and quality of runoff. Soil media with higher permeability improve the runoff reduction performance but decrease the pollution removal efficiency of the bioretention. Compared to the existing soil media design methods that are highly empirical or simplistic, the proposed design procedure relies on objective evaluation indices and quantitative analysis steps, and so may be applied to various design scenarios in stormwater management practices with improved reliability.

An evaluation index system for intellectual capital evaluation based on machine learning

Currently, there is not yet a mature evaluation index system of intellectual capital among enterprises. The lack of such a system hinders the smooth transform of capital to enterprise value. Therefore, this paper attempts to set up an effective and objective evaluation index system for intellectual capital. First, the data on intellectual capital were collected from some enterprises from the Growth Enterprise Market (GEM). Next, the original data were preprocessed into 1770 effective pieces of data. On this basis, 13 indices were selected from three dimensions (e.g. human capital, structural capital, and relationship capital) of intellectual capital, forming an evaluation index system. After that, the evaluation index system was verified with two machine learning (ML) algorithms, namely, random forest (RF), and support vector machine (SVM). The results show that our evaluation index system can optimize the intellectual capital classification of enterprises, avoiding the subjective defects in qualitative evaluation. The research results shed important new light on the decision-making and scientific management of enterprises.

Improved image defogging method based on adaptive neighborhood window size of PMS network

Aiming at the failure of prior theory of dark channel in bright area, an improved algorithm based on PMS-Net is proposed. First, the selection of atmospheric light in dark channel images under different neighborhood windows is optimized, and the atmospheric light value is accurately estimated. Then an improved transmittance estimation method is used to make the transmission map estimation of the bright area more accurate. The improved PMS-Net defogging algorithm is compared with the existing excellent algorithms, and the experimental results show that the algorithm can overcome the color distortion of existing algorithm in processing bright areas. The effectiveness of the algorithm is verified by objective evaluation indicators.

Evaluation of vehicle vibration comfort using deep learning

Traditional objective methods for vehicle vibration comfort evaluation are insufficient as they use a simple calculated objective evaluation index for evaluating the degree of discomfort. The detailed vibration contents, for example, the frequency spectrum, have been proven to be critical but are not considered comprehensively by these methods. In this study, by using deep learning, a new method is developed to assess passenger car vibration comfort based on the detailed contents of vibration signals. Four passenger cars are tested under different velocities and road types to generate an initial dataset. To overcome the limitation of dataset size, a new data augmentation method is proposed and verified for deep learning, which includes the data segmentation and resampling. Two neural networks, namely the feed-forward neural network and gated recurrent units, are used to generate two basic models, where the gated recurrent units take the time sequence characteristic of vibration data into consideration. Besides, the frequency content is considered by adding a fast Fourier transform layer to the basic models to generate two new architectures. The results suggest that the proposed data augmentation method enlarges the dataset efficiently and makes it feasible to apply deep learning to ride comfort evaluation. Also, the time sequence and frequency contents both improved the model prediction accuracy significantly, while the former manifested a greater improvement. The gated recurrent units with the fast Fourier transform layer presents the highest performance to predict the degree of comfort for new vibration scenarios. To further improve the prediction accuracy and understand the rules to apply deep learning in ride comfort evaluation, the hyperparameters of the superior architecture are tuned via parametric study. After that, the prediction loss is reduced by 76.9%. The superior performance of the final model is also demonstrated by comparing it with a traditional method.

Research on the Evaluation Index System of the Construction of Communities Suitable for Aging by the Fuzzy Delphi Method

In order to cope with the rapidly aging society and the Chinese traditional idea of old-age care, community home-based care has become a major mode of care for the aged in China, and the construction of communities suitable for the aged has also become the focus of the whole society. In order to build an objective and scientific evaluation index system of communities suitable for aging, the hierarchical structure and relative important values of the indicators for the assessment were obtained through data text rooted coding and the fuzzy Delphi expert questionnaire survey. The results prove that: the evaluation index system of communities suitable for aging consists of 4 criteria (i.e., humanistic care, public environment, health care, and economic security), 14 sub-criteria, and 48 evaluation indexes. The important values of the four criteria are “health care” (7.35), “public environment” (7.18), “humanistic care” (6.92), and “economic security” (6.87). This shows that with the aging of the elderly and the decline of physical function, health care is the most important material basis for community home care, and is also the core criterion for the construction of communities suitable for aging. Of the 48 important values of the evaluation indexes, “setting up an emergency assistance system” (7.89), “ensuring the accessibility of roads” (7.79), and “handling related affairs” (7.60) ranked the highest. This shows that the protection of the physical and mental health of the elderly and the rights and interests of the elderly are the core values of the construction of communities suitable for aging. This study believes that through constructing an evaluation index system of the communities suitable for aging, the past general constructions suitable for aging according to the standard can be further clarified through the scientific procedure of the ‘community suitable for aging construction evaluation method’, and be a reference for the academic and practical fields.

Research and Hardware Design of Image Processing Algorithm Based on FPGA

With the continuous development of computer technology and microelectronics technology, image processing technology is constantly being improved, especially real-time image processing technology has been studied by a large number of scholars. This article aims to achieve the purpose of FPGA-based fast image processing through the research of FPGA’s fast image processing algorithm. Based on the advantages of FPGA high-speed hardware computing and parallel computing, this paper proposes a design based on FPGA parallel dual template strategy, and designs a comparative experiment to verify. Obtain objective evaluation indexes of image processing results through experiments, and analyze the data results. The experimental results show that: the implementation of the mean filter algorithm based on FPGA parallel dual templates, the time required is about half that of the mean filter algorithm based on FPGA single templates. Relative to the reference Matlab software programming method to achieve 678 times performance improvement. It shows that the realization method of the mean filter algorithm based on FPGA parallel dual template is feasible. Give full play to the advantages of FPGA parallel processing data, improve the processing speed of the image algorithm module.

Evaluation of robots that signals a pedestrian using face orientation based on analysis of velocity vector fluctuation in moving trajectories

In this paper, we focus on mobile robots' behavior that people around them feel comfortable. The subjective comfort of pedestrians is usually evaluated by a questionnaire, but the high cost of assessment makes it difficult to evaluate in various situations. Towards robot control based on objective measurements of pedestrians, this paper proposes an index to evaluate the ease of walking based on objective measurements of a pedestrian's trajectory. To verify the index, we introduce a mobile robot that (1) signals it has noticed a pedestrian in front of it by moving its head upward, then (2) signals the direction of future movement by turning its head. Then the robot's behavior was evaluated by subjective impressions and objective evaluation index. We examined the relationship between subjective impressions and the objective evaluation index.

Finite element simulation on the relationship between pressure and displacement for the waist of elastic pantyhose

PurposeObjective appraisal of pressure comfort is the key point of optimal designing of clothing. The purpose of this paper is to study a new method to provide pressure comfort for the waist of elastic pantyhose through the relationship between pressure and displacement using the finite element method (FEM).Design/methodology/approachThis paper presented a simulation model of the waist cross section consisting of three parts, namely skin, soft tissue and lumbar vertebrae, respectively, according to CT scan. The finite element the model of waist cross-section was established using Mimics software. The pressure–displacement quadratic equation can be obtained using ANSYS software and fitting curves. Meanwhile, we divide the waist cross-section into 12 equal regions according to angle, and then the area shrinkage mass of the waist cross-section can be calculated, respectively.FindingsIn this research work, we got the displacement distribution trend of elastic pantyhose at the waist cross section according to the area shrinkage mass of 12 regions, and this displacement could be used as an objective evaluation index for pressure comfort. All these solutions supply a theoretical reference for optimal design of the women's elastic pantyhose.Originality/valueThe paper analyzed the relationship between pressure and displacement for the waist of elastic pantyhose using FEM, and then got the displacement distribution trend of elastic pantyhose at the waist cross section according to the area shrinkage mass of different regions. It can supply a new method to appraise pressure comfort.

Subjective Product Evaluation System Based on Kansei Engineering and Analytic Hierarchy Process

Product evaluation is very important for product improvement and development, and subjective product evaluation determines customer’s evaluation of products to some extent, so the purpose of this study is to establish a reasonable subjective product evaluation system. In this study, we comprehensively determine the evaluation indexes based on Kansei engineering (KE), establish an overall product evaluation system by using analytic hierarchy process (AHP), and establish the subjective product evaluation system by classifying the evaluation indexes in the overall product evaluation system into “objective evaluation index” and “subjective evaluation index”, removing the objective evaluation indexes, and retaining the subjective evaluation indexes. Additionally, we select some modern chairs as experimental samples to verify the reliability and validity of this subjective product evaluation system by means of questionnaires. The experimental results show that, in this subjective product evaluation system, the subjective evaluation of the product is positively correlated with the “favorite” level of the product in comprehensive evaluation, and negatively correlated with the “least favorite” level of the product in comprehensive evaluation, indicating that this subjective product evaluation system realizes a symmetry between subjective product evaluation and comprehensive product evaluation. Therefore, it can be concluded that this subjective product evaluation system based on KE and AHP proposed in this study has reliability and validity, and can be used for product evaluation to judge the popularity of products and enhance the competitiveness of products.

A novel multi-modal medical image fusion algorithm

In order to improve the contrast of image fusion and highlight the unique characteristics of medical images, a multi-modal medical image fusion algorithm in the framework of non-subsampled contourlet transform (NSCT) is proposed in this paper. Firstly, the computed tomography images and magnetic resonance image are decomposed into low- and high-frequency sub-bands through the NSCT of multi-scale geometric transformation; secondly, for the low-frequency sub-band, the local area standard deviation method is selected or the fusion, while for the high-frequency sub-band, an adaptive pulse coupling neural network model is constructed and the fusion rules are set by the cumulative ignition times of iterative operation in the network; finally, the fusion image is obtained through image reconstruction. Experimental results show that the fusion results of the algorithm in this paper can improve the image fusion quality significantly and it has certain advantages in both visual effects and objective evaluation indexes, which provides a more reliable basis for clinical diagnosis and treatment of diseases.

Attention‐based end‐to‐end image defogging network

Aiming at the problem that the traditional prior information-based defogging algorithm fails in some special scenarios, an end-to-end convolutional defogging network based on attention mechanism is proposed. The network consists of two modules: parameter estimation and image restoration. First, multi-scale convolution is used to extract image feature information. Residual network and skip connection methods are used to improve the utilisation rate of shallow network feature information. Secondly, the channel domain attention is used to add weight to the feature image input from the previous network and select useful feature information. Finally, the atmospheric visibility model is combined to achieve image visibility restoration. The experimental results show that the proposed algorithm can effectively improve the visibility of the image and the restoration effect is natural. The objective evaluation index of the benchmark datasets also shows the effectiveness of the algorithm.

ANP-based fuzzy comprehensive evaluation system for electric power environment

In order to evaluate the environmental situation of electric power business in Lishui City, a relatively objective evaluation index system was established by using literature review method and Delphi method. The network analytic network process (ANP) and fuzzy comprehensive evaluation method are combined to comprehensively evaluate the electric power business environment in Lishui. Firstly, the weight of each evaluation index is determined by ANP method, which overcomes the defect that the analytic hierarchy process (AHP) can not reflect the mutual influence relationship between indicators. Then, the ANP-based fuzzy comprehensive evaluation model is constructed to evaluate the uncertain electric power business environment and calculate with the help of Super Decisions software. The results show that the electricity business environment in Lishui is better. Finally, some measures to improve the electricity business environment were proposed based on the corresponding indicators.

Objective assessment of nasal function changes after revision endoscopic sinus surgery in patients with recurrent sinusitis

Objective:To investigate the objective evaluation system of nasal function changes in patients with recurrent sinusitis after revision endoscopic sinus surgery. Method:Sixty patients with recurrent sinusitis underwent revision endoscopic sinus surgery. The Lund-Kennedy endoscopic sinus score and Lund-Mackay paranasal sinus CT score were performed before surgery, 3, 6 and 12 months after surgery. The saccharin test was used to evaluate the nasal mucosa ciliary moving function（nasal mucosal ciliary motor velocity） before surgery, 6 and 12 months after surgery. The correlation of objective evaluation indexes before and after surgery was analyzed. Result:The Lund-Kennedy endoscopic sinus score and Lund-Mackay paranasal sinus CT score of patients 3, 6, 12 months after surgery were lower than those before surgery（P<0.05）; The Lund-Kennedy endoscopic sinus score and Lund-Mackay paranasal sinus CT score of patients 6 and 12 months after surgery were lower than those 3 months after surgery（P<0.05）; The Lund-Kennedy endoscopic sinus score and Lund-Mackay paranasal sinus CT score of patients 12 months after surgery were lower than those 6 months after surgery（P<0.05）. The nasal mucosal ciliary motor velocity of patients 6 and 12 months after surgery was lower than that before surgery（P<0.05）; The nasal mucosal ciliary motor velocity of patients 12 months after surgery was lower than that 6 months after surgery（P<0.05）. The Lund-Kennedy endoscopic sinus score was positively correlated with Lund-Mackay paranasal sinus CT score in patients before surgery, 3, 6 and 12 months after surgery（P<0.05）; The Lund-Kennedy endoscopic sinus score was negatively correlated with nasal mucosal ciliary motor velocity in patients before surgery, 6 and 12 months after surgery（P<0.05）; The Lund-Mackay paranasal sinus CT score was negatively correlated with nasal mucosal ciliary motor velocity in patients before surgery, 6 and 12 months after surgery（P<0.05）. Conclusion:There is good correlation among the Lund-Kennedy endoscopic sinus score, the Lund-Mackay paranasal sinus CT score and the nasal mucosal ciliary motor velocity in patients with recurrent sinusitis before and after revision endoscopic sinus surgery. This study establishes an objective evaluation system for evaluating the changes in nasal function with these three indexes as the main body.