International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 11 Issue: 07 | July 2024
www.irjet.net
p-ISSN: 2395-0072
The framework integrates contrast enhancement and edge detection for effective segmentation of low-quality images Alaa N. Alanssari Lecturer Professor, Department of Mathematic, Faculty of Basic Education, University of Kufa, Najaf, Iraq ---------------------------------------------------------------------***--------------------------------------------------------------------
Abstract - The meeting of noise and low contrast typically
activity is to isolate the main objects in the foreground from their background, which holds significance in areas such as medical imaging and remote sensing, along with object recognition systems. Successful segmentation paves way for valuable feature information that should ideally be extracted for classification and image analysis; however, achieving high-quality delineation from noisy or low contrast images is a Herculean task. An impasse difficulty many a researcher faces while working on this topic hindering their progress.
amplifies the difficulty of segmenting images by their dual nature. However, we demonstrate an unconventional approach: attempting to achieve the utmost degree of accuracy in segmenting via a hybridization of multiple methods into a single flow. Let's begin with CLAHE, which takes charge of local contrast enhancement; followed by Gaussian smoothing, an actor that possesses dual capabilities first, they retain information regarding the edge of the object, and second, they eliminate noise that is not tolerated. Then Canny edge detection: this method is used to differentiate strong pixels from thin, weak ones, the method is based on the gradient of the pixels.
Common approaches to segmenting, such as thresholding, region growing, and edge detection, typically require the presence of a clear distinction between regions or a uniformity within the regions. While these methods are effective in certain situations, they are greatly limited by noise or low contrast that makes them ineffective [1]. Similarly, traditional approaches that use segmentation are unsuccessful in situations like: remote sensing, where a common quality sensor is obstructed by environmental areas that are standard, but the contrast between these areas is low, this prevents the identification of features [2].
We don't conclude here. After the Canny procedure, the role of the morphological procedure is significant. They improve the map of the edge by taking care of the small regions of noise and closing the gaps (these are not welcome). This is the methodology we follow! Through experimental results that have already been successful, it has surpassed other methods in terms of accuracy when dealing with images that are of low quality, this is an example of innovation! This paints a bright picture of how effective our fusion strategy is likely to be; it represents a significant advancement since we address specific details that are embedded in noise (commonly encountered in tasks like object recognition or image analysis) at specific locations. The capacity of our methodology to deal with different image situations and the effectiveness of the many types of noise it can address can be considered as a significant advantage an advantage that can be utilized to enhance the effort towards improving the delineation of images (hence this process in challenging environments). To end: we're transferring a superior product.
A new method of segmenting images is under development. A more advanced one. not your typical highquality images that other methods have an easier time dealing with because of their limitations. One innovation that is particularly noteworthy is adaptive filtering: a method that adjusts its parameters based on the specifics of the image at hand. When combined with cutting-edge edge detection technology that is highly resilient to noise, we have a solution that can recognize features regardless of what others are doing. Among the exhibited innovative approaches is Contrast Limited Adaptive Histogram equalization (CLAHE), which is renowned for its emphasis on enhancing local contrast enhancement. Disparate lighting conditions that are adverse to the standard technique often lead to features that are more apparent than other techniques would have missed. These conditions are typically caused by information that is background. This is an example of image segmentation that is redefined, every new method provides a path to more complex and durable solutions to real-world problems, but which are still undisclosed... One additional
Key
words: Segmentation, Low-Quality Images, Contrast Enhancement, Edge Detection, Morphological Operations. 1.INTRODUCTION Splitting an image into meaningful pieces can be termed as the most important aspect of handling images this is what we refer to as segmentation. The primary goal of this
© 2024, IRJET
|
Impact Factor value: 8.226
|
ISO 9001:2008 Certified Journal
|
Page 613