Abstract:Image enhancement is a fundamental prerequisite for every computer vision program that intends to process an image further. When applied to practically undetectable photos, one of the most common limitations of most existing approaches is the loss of color information throughout the improvement process. In this work, a method based on the enhancement of base as well as detail images using the concept of global and local enhancement for highly degraded images has been proposed which improves the highly degraded image along with preserving its color as well as naturalness. In this novel method, a global–local technique is proposed that breaks the image into smoother and sharper regions called base and detail images. The base image consists of the smoother regions in the input image whereas the detail image contains the sharp edges. The global approach is applied for improving the base image and the local technique is applied for the enhancement of the detail image, containing the sharper edges. The base and detail images are estimated using the median and weighted least squares (WLS) filters respectively. The base image is enhanced using the global approach using the modified form of AGC based on the cumulative histogram. The value of gamma is derived from the image parameters, which makes the proposed method adaptive and applicable to a wide range of images with different contrast degradations. The detail image is enhanced using the newly introduced parameter RoE, which ensures that the enhancement of the detail image is in fine tune with the base image. The enhanced base and detail images are combined and scaled to bring the intensity levels to the permitted range. Finally, mean adjustment is applied to develop the final improved image. The approach improves visual contrast while preserving naturalness. The simulation results on typical datasets show that the suggested technique outperforms numerous state-of-the-art as well as traditional algorithms for extremely degraded images.

Globally and locally tuned filtering structure for high contrast intensity degradation

Microscopic image detail enhancement based on local window and extremum

Highly Robust Real-Time Image Enhancement

Low-light Image Enhancement with Strong Light Weakening and Bright Halo Suppressing.

Color Retinal Image Enhancement Based on Luminosity and Contrast Adjustment

Computer-aided Diagnosis Based on Enhancement of Degraded Fundus Photographs.

Image contrast enhancement with brightness preservation using an optimal gamma and logarithmic approach

Global–local image enhancement with contrast improvement based on weighted least squares

Contrast Enhancement Via Dual Graph Total Variation-Based Image Decomposition.

Optimized multi-scale framework for image enhancement using spatial information-based histogram equalization

A Low-Light Image Enhancement Method for Both Denoising and Contrast Enlarging.

Underwater Image Enhancement by Attenuated Color Channel Correction and Detail Preserved Contrast Enhancement

Noise reduction deep CNN-based retinal fundus image enhancement using recursive histogram

Effective solution for underwater image enhancement

Amalgamation of ROAD-TGM and progressive PCA using performance booster method for detail persevering image denoising

Infrared Image Enhancement Based on Adaptive Guided Filter and Global–Local Mapping

Generation of High Dynamic Range Illumination from a Single Image for the Enhancement of Undesirably Illuminated Images

Low-light and hazy image enhancement using retinex theory and wavelet transform fusion

Brightness Aware Pixel Stretching for Perceptually Invisible Images Using Wavelet Approximation Balancing

Effective Enhancement Method of Low-Light-level Images Based on the Guided Filter and Multi-Scale Fusion.

Enhancement and Noise Suppression of Single Low-Light Grayscale Images