Image Enhancement By Detecting And Removing Rain In Digital Images

Vinoad, Ganesan (2015) Image Enhancement By Detecting And Removing Rain In Digital Images. Project Report. UTeM, Melaka, Malaysia. (Submitted)

Text (24 Pages) Image Enhancement By Detecting And Removing Rain In Digital Images 24 Pages.pdf - Submitted Version Download (415kB)

Abstract

Different climate conditions, for example, rain, snow, or cloudiness will result in intricate visual impacts on spatial or fleeting spaces in images or videos. Such effects may significantly degrade the performances of outdoor vision systems. Poor perceivability debases perceptual quality of image and execution of the machine vision calculations, for example, surveillance, detection, recognition tracking, and navigation. The raindrops debase the execution of open air vision framework, and it brings troubles for items discovery and examination in a solitary picture. Identification and removal of rain in an image is a troublesome and critical issue because of the unpredictability of rain and its negative consequences for image. Removal of rain from images as a pre-processing builds the correctness of these computer vision calculations. A peculiarity point locator can fizzle if pictures have low perceivability. On the off chance that rain is removed and picture is improved, then investment point recognition can work with higher exactness. In this thesis, the bilateral filter and Guided filter have been applied for rain removal and image enhancement. The rain drops in digital images successfully detected by using Hough transform. The detected rain drops removed and the image enhance by using guided filter. The similar procedures done with the bilinear filter in order for comparison. As expected, the resulting images of bilateral filtering show little details and still contain rain. The Guided filter performed marginally better than the bilateral filter. More rain drops was removed by this method. The results from the guided filter shows no rain drops and smooth edges. Guided filter able to remove the rain and enhance the removed image to perform the best quality image.

Actions (login required)

View Item