Aysun Taşyapı Çelebi

Visual enhancement of underwater images using Empirical Mode Decomposition

Most underwater vehicles are nowadays equipped with vision sensors. However, it is very likely that underwater images captured using optic cameras have poor visual quality due to lighting conditions in real-life applications. In such cases it is useful to apply image enhancement methods to increase visual quality of the images as well as enhance interpretability and visibility. In this paper, an Empirical Mode Decomposition (EMD) based underwater image enhancement algorithm is presented for this purpose. In the proposed approach, initially each spectral component of an underwater image is decomposed into Intrinsic Mode Functions (IMFs) using EMD. Then the enhanced image is constructed by combining the IMFs of spectral channels with different weights in order to obtain an enhanced image with increased visual quality. The weight estimation process is carried out automatically using a genetic algorithm that computes the weights of IMFs so as to optimize the sum of the entropy and average gradient of the reconstructed image. It is shown that the proposed approach provides superior results compared to conventional methods such as contrast stretching and histogram equalizing.

Fuzzy fusion based high dynamic range imaging using adaptive histogram separation

In this work, a high dynamic range (HDR) image generation method using a single input image is presented. The proposed approach generates over- and under-exposed images by making use of a novel adaptive histogram separation scheme. Thus, it becomes possible to eliminate ghosting effects which generally occur when several input image containing camera/object motion are utilized in HDR imaging. Additionally, it is proposed to utilize a fuzzy logic based approach at the fusion stage which takes visibility of the inputs pixels into account. Since the proposed approach is computationally light-weight, it is possible to implement it on mobile devices such as smart phones and compact cameras. Experimental results show that the proposed approach is able to provide ghost-free and improved HDR performance compared to the existing methods.

Empirical mode decomposition based visual enhancement of underwater images

Most underwater vehicles are nowadays equipped with vision sensors. However, underwater images captured using optic cameras can be of poor quality due to lighting conditions underwater. In such cases it is necessary to apply image enhancement methods to underwater images in order to enhance visual quality as well as interpretability. In this paper, an Empirical Mode Decomposition (EMD) based image enhancement algorithm is applied to underwater images for this purpose. EMD has been shown to be particularly suitable for non-linear and non-stationary signals in the literature, and therefore provides very useful in real life applications. In the approach presented in this paper, initially each R, G and B channel of the color underwater image is separately decomposed into Intrinsic Mode Functions (IMFs) using EMD. Then, the enhanced image is constructed by combining the IMFs of each channel with different weights, so as to obtain a new image with increased visual quality. It is shown that the proposed approach provides superior results compared to conventional image enhancement methods such as contrast stretching.

Enhancement of fog degraded images using empirical mode decomposition

Images can have poor visibility, contrast and colors in foggy weather conditions. Therefore it is required to enhance visual quality of the fog-degraded images. In this paper we present a new method based on an Empirical Mode Decomposition (EMD) for fog-degraded image enhancement. Initially each spectral component of the fog-degraded image is decomposed into Intrinsic Mode Functions (IMFs) using EMD. Then the enhanced image is constructed by combining the IMFs of spectral channels with optimum weights in order to obtain an enhanced image with increased visual quality. The optimal weight estimation process is carried out automatically using genetic algorithm. Eventually, image enhancement is completed performing color correction followed by a de-quantization.

Visual enhancement of underwater images using Empirical Mode Decomposition and wavelet denoising

In recent years, most underwater vehicles are equipped with optical cameras to capture underwater images. But underwater images acquired using optic cameras have poor visual quality due to propagation of properties of light in water. So it is useful to apply image enhancement methods to increase visual quality of the images as well as enhance interpretability and visibility. In this paper, an Empirical Mode Decomposition (EMD) based underwater image enhancement algorithm is presented for this purpose. In the proposed approach, initially each color channel (R, G, B) of an underwater image is decomposed into Intrinsic Mode Functions (IMFs) using EMD. The first IMF of each component is applied to wavelet denoising. Because this IMF includes all local high spatial frequency components. Then the enhanced image is constructed by combining the IMFs of spectral channels with different weights in order to obtain an enhanced image with increased visual quality. The weight estimation process is carried out automatically using a genetic algorithm that computes the weights of IMFs so as to optimize the sum of the entropy and average gradient of the reconstructed image.

Mine detection in side scan sonar images using Markov Random Fields with brightness compensation

In this paper a new mine detection algorithm for side scan sonar images is proposed. In the proposed method, an illumination normalization algorithm to compensate for illumination variations based on Discrete Cosines Transform is applied initially to the sonar images. Then, segmentation is carried out using Markov Random Fields and potential mines are detected.

Target detection in sonar images using Empirical Mode Decomposition and morphology

In this paper to a new target detection algorithm for sonar images based on segmention using Empirical Mode Decomposition (EMD) and Morphological Operations is proposed. In the proposed approach morphological operations are applied after EMD to provide higher detection accuracy. Experimental results show that the EMD based approach proposed in this paper improves target detection accuracy. The target detection performance is increased compared to utilizing morphological operations only, as available in the literature.

Integer 1-bit transform method and its hardware architecture for low-complexity block-based motion estimation

The need for coding efficiency is being increased with applications in which high resolution video processing is being performed. Power consumption and memory are important constraints for devices capable of recording and transmitting video, such as ultra-high definition televisions (UHD TVs), cameras, smartphones. In video encoders, motion estimation is the process which utilizes the most complex tasks and consumes most of the power. Therefore, low complexity motion estimation methods have been developed which can provide efficient hardware architectures. In this work, a novel integer 1-bit transform method is proposed. Additionally, proposed method and multiplication-free one bit transform method are compared by taking hardware cost which originates from binarization process into account and experimental results are presented.

Local binary pattern method and its hardware architecture for low-complexity motion estimation

In video encoding, most of the computational load originates from motion estimation. Low complexity motion estimation methods have been developed in the literature in order to reduce the complexity of motion estimation which consumes most of the total power. By using the proposed local binary pattern based method at the binarization stage, pixels are represented by using only 1-bit instead of 8-bits. In this work, local binary pattern method is utilized as a low complexity motion estimation method and a hardware architecture is proposed for both binarization and matching stage.

Implementation of 1 -bit transform based motion estimation to HEVC encoder

In this work, one-bit transform (1BT) based motion estimation method is applied to HEVC encoder which is the most recent video coding standard. The most significant portion of the computational load of the HEVC originates from the motion estimation. Thus, many approaches are presented in literature in order to reduce this computational load. 1BT based ME approach denotes input image frames in low-bit depth form so that it is not only suitable for hardware implementation but also reduces computational load of whole process since it uses EX-OR based matching criterion. In this work, 1BT based ME is integrated into HEVC encoder and its performance is assessed. Experimental results show that the 1BT based ME approach has a neglectable performance loss compared conventional HEVC motion estimation approach.