Channel (digital image)

About: Channel (digital image) is a research topic. Over the lifetime, 7211 publications have been published within this topic receiving 69974 citations.

Deep Learning Based Single Image Dehazing

Abstract: This paper proposes a novel approach to remove haze degradations in RGB images using a stacked conditional Generative Adversarial Network (GAN). It employs a triplet of GAN to remove the haze on each color channel independently. A multiple loss functions scheme, applied over a conditional probabilistic model, is proposed. The proposed GAN architecture learns to remove the haze, using as conditioned entrance, the images with haze from which the clear images will be obtained. Such formulation ensures a fast model training convergence and a homogeneous model generalization. Experiments showed that the proposed method generates high-quality clear images.

Coloring night-vision imagery with statistical properties of natural colors by using image segmentation and histogram matching

Abstract: A natural color mapping method has been previously proposed that matches the statistical properties (mean and standard deviation) of night-vision (NV) imagery to those of a daylight color image (manually selected as the "target" color distribution). Thus the rendered NV image appears to resemble the target image in terms of colors. However, in this prior method the colored NV image may appear unnatural if the target image's "global" color statistics are too different from that of the night vision scene (e.g., it would appear to have too much green if much more foliage was contained in the target image). Consequently, a new "local coloring" method is presented in the current paper, and functions to render the NV image segment-by-segment by using a histogram matching technique. Specifically, a false-color image (source image) is formed by assigning multi-band NV images to three RGB (red, green and blue) channels. A nonlinear diffusion filter is then applied to the false-colored image to reduce the number of colors. The final grayscale image segments are obtained by using clustering and merging techniques. The statistical matching procedure is merged with the histogram matching procedure to assure that the source image more closely resembles the target image with respect to color. Instead of using a single target color image, the mean, standard deviation and histogram distribution of a set of natural scene images are used as the target color properties for each color scheme. Corresponding to the source region segments, the target color schemes are grouped by their scene contents (or colors) such as green plants, roads, ground/earth. In our experiments, five pairs of night-vision images were initially analyzed, and the images that were colored (segment-by-segment) by the proposed "local coloring" method are shown to be much more natural, realistic, and colorful when compared with those produced by the "global-coloring" method.

A fast color image enhancement algorithm based on Max Intensity Channel

Abstract: In this paper, we extend image enhancement techniques based on the retinex theory imitating human visual perception of scenes containing high illumination variations. This extension achieves simultaneous dynamic range modification, color consistency, and lightness rendition without multi-scale Gaussian filtering which has a certain halo effect. The reflection component is analyzed based on the illumination and reflection imaging model. A new prior named Max Intensity Channel (MIC) is implemented assuming that the reflections of some points in the scene are very high in at least one color channel. Using this prior, the illumination of the scene is obtained directly by performing a gray-scale closing operation and a fast cross-bilateral filtering on the MIC of the input color image. Consequently, the reflection component of each RGB color channel can be determined from the illumination and reflection imaging model. The proposed algorithm estimates the illumination component which is relatively smooth and maintains the edge details in different regions. A satisfactory color rendition is achieved for a class of images that do not satisfy the gray-world assumption implicit to the theoretical foundation of the retinex. Experiments are carried out to compare the new method with several spatial and transform domain methods. Our results indicate that the new method is superior in enhancement applications, improves computation speed, and performs well for images with high illumination variations than other methods. Further comparisons of images from National Aeronautics and Space Administration and a wearable camera eButton have shown a high performance of the new method with better color restoration and preservation of image details.

MLOD: A multi-view 3D object detection based on robust feature fusion method

Abstract: This paper presents Multi-view Labelling Object Detector (MLOD). The detector takes an RGB image and a LIDAR point cloud as input and follows the two-stage object detection framework [1] [2]. A Region Proposal Network (RPN) generates 3D proposals in a Bird’s Eye View (BEV) projection of the point cloud. The second stage projects the 3D proposal bounding boxes to the image and BEV feature maps and sends the corresponding map crops to a detection header for classification and bounding-box regression. Unlike other multi-view based methods, the cropped image features are not directly fed to the detection header, but masked by the depth information to filter out parts outside 3D bounding boxes. The fusion of image and BEV features is challenging, as they are derived from different perspectives. We introduce a novel detection header, which provides detection results not just from fusion layer, but also from each sensor channel. Hence the object detector can be trained on data labelled in different views to avoid the degeneration of feature extractors. MLOD achieves state-of-the-art performance on the KITTI 3D object detection benchmark. Most importantly, the evaluation shows that the new header architecture is effective in preventing image feature extractor degeneration.

Channel Attention Residual U-Net for Retinal Vessel Segmentation

Abstract: Retinal vessel segmentation is a vital step for the diagnosis of many early eye-related diseases. In this work, we propose a new deep learning model, namely Channel Attention Residual U-Net (CAR-UNet), to accurately segment retinal vascular and non-vascular pixels. In this model, we introduced a novel Modified Efficient Channel Attention (MECA) to enhance the discriminative ability of the network by considering the interdependence between feature maps. On the one hand, we apply MECA to the "skip connections" in the traditional U-shaped networks, instead of simply copying the feature maps of the contracting path to the corresponding expansive path. On the other hand, we propose a Channel Attention Double Residual Block (CADRB), which integrates MECA into a residual structure as a core structure to construct the proposed CAR-UNet. The results show that our proposed CAR-UNet has reached the state-of-the-art performance on three publicly available retinal vessel datasets: DRIVE, CHASE DB1 and STARE.