Due to the strong reflection property of materials with smooth surfaces like ceramic and metal, it will cause saturation and the highlight phenomenon in the image when taking pictures of those materials. In order to solve this problem, a new algorithm which is based on reflection component separation (RCS) and priority region filling theory is designed. Firstly, the specular pixels in the image are found by comparing the pixel parameters. Then, the reflection components are separated and processed. However, for ceramic, metal and other objects with strong specular highlight, RCS theory will change color information of highlight pixels due to larger specular reflection component. In this situation, priority region filling theory was used to restore the color information. Finally, we implement 3D experiments on objects with strong reflecting surfaces like ceramic plate, ceramic bottle, marble pot and yellow plate. Experimental results show that, with the proposed method, the highlight caused by the strong reflecting surface can be well suppressed. The highlight pixel number of ceramic plate, ceramic bottle, marble pot and yellow plate, is decreased by 43.8 times, 41.4 times, 33.0 times, and 10.1 times. Three-dimensional reconstruction results show that highlight areas were significantly reduced.

Hyperspectral image (HSI) classification methods based on three-dimensional convolutional neural network (3DCNN) have problems of overfitting the in-sample training process and difficulty in highlighting the role of discriminant features, which reduce the classification accuracy. To solve the above problems, an HSI classification method based on M-3DCNN-Attention is proposed. First, the Mixup algorithm is used to construct HSI virtual samples to expand the original data set. The sample size of the expanded data set is twice that of the original data set, which greatly alleviates the overfitting phenomenon caused by the small sample of HSI. Second, the structure of 3DCNN is improved. A convolutional block attention module (CBAM) is added between each 3D convolutional layer and ReLU layer, and a total of three CBAMs are used so as to highlight the discriminant features in spectral and spatial dimensions of HSI and suppress the nondiscriminant features. Finally, the spectral–spatial features are transferred to the Softmax classifier to obtain the final classification results. The comparative experiments are conducted on three hyperspectral data sets (Indian Pines, University of PaviaU, and Salinas), and the overall accuracy of M-3DCNN-Attention is 99.90%, 99.93%, and 99.36%, respectively, which is better than the comparative methods.

Medical images play an important role in the hospital diagnosis and treatment, which include a lot of valuable medical information. Manually annotated viewing is obviously not effective in managing large amounts of medical imaging data. Hence it is an important task to establish an efficient and accurate medical image retrieval system. In this paper, a medical image retrieval approach based on Hausdorff distance combining Tamura texture features and wavelet transform algorithm is proposed. The combination of Tamura texture features and wavelet transform features can extract the texture features of medical images more effectively, and Hausdorff distance can reflect the overall similarity of medical image feature set. In this paper, 6 group experiments of brain MRI database and the lung CT database were conducted separately. Experiments show that the proposed approach has higher accuracy than a single feature texture algorithm and is also higher than the approach of Tamura texture features and wavelet transform features combined with Euclidean distance.

With the wave of artificial intelligence and deep learning sweeping the world, there are many algorithms based on deep learning for image defog research. However, there is still serious color distortion, contrast reduction, incomplete fog removal, and other problems. To solve these problems, this paper proposes an improved image defogging network based on the traditional cycle-consistent adversarial network. We add the self-attention module and atrous convolution multi-scale feature fusion module on the basis of the traditional CycleGAN network to enhance the feature extraction capability of the network. The perceptual loss function is introduced into the loss function of the model to enhance the texture sense of the generated image. Finally, by comparing several typical defogging algorithms, the superiority of the defogging model proposed in this paper is proved qualitatively and quantitatively. Among them, on the indoor synthetic data set, the Peak Signal-to-Noise Ratio (PSNR) and Structural Similarity Index Measurement (SSIM) of the network designed by us can reach 23.22 and 0.8809, respectively. On the outdoor synthetic data set, the PSNR and SSIM of our designed network can be as high as 25.72 and 0.8859, respectively. On the real data set, the PSNR and SSIM of our designed network can reach 21.02 and 0.8166, respectively. It is proved that the defogging network in this paper has good practicability and universality.

The purpose of image registration is to find the symmetry between the reference image and the image to be registered. In order to improve the registration effect of unmanned aerial vehicle (UAV) remote sensing imagery with a special texture background, this paper proposes an improved scale-invariant feature transform (SIFT) algorithm by combining image color and exposure information based on adaptive quantization strategy (AQCE-SIFT). By using the color and exposure information of the image, this method can enhance the contrast between the textures of the image with a special texture background, which allows easier feature extraction. The algorithm descriptor was constructed through an adaptive quantization strategy, so that remote sensing images with large geometric distortion or affine changes have a higher correct matching rate during registration. The experimental results showed that the AQCE-SIFT algorithm proposed in this paper was more reasonable in the distribution of the extracted feature points compared with the traditional SIFT algorithm. In the case of 0 degree, 30 degree, and 60 degree image geometric distortion, when the remote sensing image had a texture scarcity region, the number of matching points increased by 21.3%, 45.5%, and 28.6%, respectively and the correct matching rate increased by 0%, 6.0%, and 52.4%, respectively. When the remote sensing image had a large number of similar repetitive regions of texture, the number of matching points increased by 30.4%, 30.9%, and −11.1%, respectively and the correct matching rate increased by 1.2%, 0.8%, and 20.8% respectively. When processing remote sensing images with special texture backgrounds, the AQCE-SIFT algorithm also has more advantages than the existing common algorithms such as color SIFT (CSIFT), gradient location and orientation histogram (GLOH), and speeded-up robust features (SURF) in searching for the symmetry of features between images.

Safety helmet plays a vital role in protecting worker’s head in dangerous working environment. During the inspection of safety helmets by means of video automatic monitoring, due to the complex factory environment and large area, the safety helmets worn by workers in the foreground are not easy to be detected, resulting in the problem of safety helmet leakage. To solve this problem, a helmet detection method combining multi-feature fusion and support vector machine (SVM) is proposed to improve the helmet recognition rate. First of all, the visual background difference algorithm was used to detect workers, and the initial positioning of the helmet was determined by the proportional relationship between the head and the whole body. Secondly, this paper uses the principal component analysis method (PCA algorithm) to reduce the dimensionality of the feature vector, cascades the two feature vectors after the dimensionality reduction with the center of gravity, and use the SVM model based on Bayesian optimization to identify the helmet. Finally, a method combining and Meanshift algorithm of multi-feature fusion and Kalman filter is proposed to track the detected helmet. The average recognition rate of multiple experiments is 90.03%. So the helmet tracking algorithm combined with Kalman filter and Meanshift of multi-feature fusion improves the helmet tracking accuracy.