Deep learning-based histopathological segmentation for whole slide images of colorectal cancer in a compressed domain

Automatic pattern recognition using deep learning techniques has become increasingly important. Unfortunately, due to limited system memory, general preprocessing methods for high-resolution images in the spatial domain can lose important data information such as high-frequency information and the region of interest. To overcome these limitations, we propose an image segmentation approach in the compressed domain based on principal component analysis (PCA) and discrete wavelet transform (DWT). After inference for each tile using neural networks, a whole prediction image was reconstructed by wavelet weighted ensemble (WWE) based on inverse discrete wavelet transform (IDWT). The training and validation were performed using 351 colorectal biopsy specimens, which were pathologically confirmed by two pathologists. For 39 test datasets, the average Dice score, the pixel accuracy, and the Jaccard score were 0.804 ± 0.125, 0.957 ± 0.025, and 0.690 ± 0.174, respectively. We can train the networks for the high-resolution image with the large region of interest compared to the result in the low-resolution and the small region of interest in the spatial domain. The average Dice score, pixel accuracy, and Jaccard score are significantly increased by 2.7%, 0.9%, and 2.7%, respectively. We believe that our approach has great potential for accurate diagnosis.

Deep Learning-Based Object Tracking via Compressed Domain Residual Frames

The extensive rise of high-definition CCTV camera footage has stimulated both the data compression and the data analysis research fields. The increased awareness of citizens to the vulnerability of their private information, creates a third challenge for the video surveillance community that also has to encompass privacy protection. In this paper, we aim to tackle those needs by proposing a deep learning-based object tracking solution via compressed domain residual frames. The goal is to be able to provide a public and privacy-friendly image representation for data analysis. In this work, we explore a scenario where the tracking is achieved directly on a restricted part of the information extracted from the compressed domain. We utilize exclusively the residual frames already generated by the video compression codec to train and test our network. This very compact representation also acts as an information filter, which limits the amount of private information leakage in a video stream. We manage to show that using residual frames for deep learning-based object tracking can be just as effective as using classical decoded frames. More precisely, the use of residual frames is particularly beneficial in simple video surveillance scenarios with non-overlapping and continuous traffic.

Design of the Poster Image System Based on Human Vision

At present, the human visual perception system is the most effective, accurate, and fast image processing system in the world. This is because human eyes have some special visual features, but the features closely related to image enhancement include color constancy and brightness constancy. This paper presents a new image enhancement framework and computational model which can better simulate human visual features. It is based on the analysis of color constancy and luminance constancy and Retinex theory. And, this is a new image enhancement method in the compressed domain based on Retinex theory. In Retinex theory, DCT coefficients consist of incident components (DC coefficients) and reflection components (AC coefficients). By adjusting the dynamic range of DC coefficients, carefully adjusting AC coefficients, and using the threshold method for block suppression, the compressed domain image can be enhanced. On the basis of Retinex theory, the incident light and reflected light components are considered synthetically, the dynamic range (DC coefficient) of the incident light component and the details of the reflected light component (AC coefficient) are adjusted, and then the incident light component is reexamined. Moreover, it achieves a better image enhancement effect and avoids the blocking effect.