The rapid development of the Industrial Internet of Things and artificial intelligence has brought new pressures to the compression and storage of image information. Thus, efficient image compression approaches are important in the fourth industrial revolution (Industry 4.0), which can save storage resources, accelerate transmission speed, and improve imaging fidelity. In this article, a new approach of image compression based on Haar-V functions (ICHV) is investigated, which has good performance in industrial image compression. The Haar-V functions form a discontinuous orthogonal function system composed of piecewise polynomials. The construction and properties of Haar-V functions are introduced at first. Then, 2-D discrete orthogonal matrices of Haar-V functions and the corresponding image transform method are presented successfully. Finally, a fast computation of ICHV is proposed to enhance the computation speed of image compression. The experimental results indicate that compared with related algorithms, the new approach can effectively improve the quality of compressed images and consume less computational cost.

Digital image feature recognition is significant to industrial information applications, such as bioengineering, medical diagnosis, and machinery industry. In order to supply an effective and reasonable technology of the severity assessment mission of coronavirus disease (COVID-19), in this article, we propose a new method that identifies rich features of lung infections from a chest computed tomography (CT) image, and then assesses the severity of COVID-19 based on the extracted features. First, in a chest CT image, the lung contours are corrected for the segmentation of bilateral lungs. Then, the lung contours and areas are obtained from the lung regions. Next, the coarseness, contrast, roughness, and entropy texture features are extracted to confirm the COVID-19 infected regions, and then the lesion contours are extracted from the infected regions. Finally, the texture features and V-descriptors are fused as an assessment descriptor for the COVID-19 severity estimation. In the experiments, we show the feature extraction and lung lesion segmentation results based on some typical COVID-19 infected CT images. In the lesion contour reconstruction experiments, the performance of V-descriptors is compared with some different methods, and various feature scores indicate that the proposed assessment descriptor reflects the infected ratio and the density feature of the lesions well, which can estimate the severity of COVID-19 infection more accurately.

Since digital images are able to be encoded through the luminance-bandwidth-chrominance (YUV) mode, and the contribution of luminance information is greater than that of chrominance information for human visual perception, it can be inferred that the appropriate reduction of chrominance information in convolutional neural network does not disturb image object recognition. In this paper, we propose a new multi-scale non-uniform convolution called YuvConv, wherein the output feature map of the convolutional layer is regarded as an image. First, the output channels in the new convolution are divided into three kinds of components: Y, U, and V tensors. Then, the tensor Y is used to process luminance information, which is high-resolution and occupies more output channels. Next, the tensors U and V are low-resolution and use fewer channels to process chrominance information. Finally, the adjacent tensors (Y-U, Y-U-V, and U-V) are fused as the output of YuvConv. Experimental results indicate that the use of the YuvConv instead of the standard convolution can improve the performance of deep learning tasks, and it can also reduce memory consumption and computation cost.

The partition algorithm as a digital image processing technique is significant to many applications, such as data encryption, image denoising, and 3-D reconstruction. In order to achieve well partition that can availably reduce the distortion phenomenon, a novel approach named image adaptive triangular partition (IATP) is proposed, which considers the grayscale distribution of the image and removes the shared edges between the adjacent triangles in the partition mesh. The least-squares method is used to fit the sampled position-associated gray value of the image to determine whether further partition should be performed, that is, if the sum of squared residuals is bigger than the preselected control value, the current area will be divided into four separated sub-triangles by using the self-similar method, and then preparing the next fitting on each of them in recursion; otherwise, the terminal operation is reached. When the recursive partition of the image is done, the triangular partition mesh with the quaternary notations is obtained. The experimental results demonstrate that the performance of the IATP algorithm proposed in this paper is better than the existing state-of-the-art nonuniform partitions, and it solves the redundant coding problem and reduces the image quality losses. In addition, two applications-image steganography and information encryption-are selected to verify that the proposed algorithm has good feasibility and robustness.