Image based fog detection in vehicles

Abstract

Abstract-Modern vehicles are equipped with many cameras and their use in many practical applications is extensive. Detecting the presence of fog from images of a camera mounted in vehicles is a very challenging task with the potential to be used in many practical applications. Approaches introduced until now analyze properties of local objects in the image like lane markings, traffic signs, back lights of vehicles in front or head lights of approaching vehicles. By contrast to all these related works we propose to use image descriptors and a classification procedure in order to distinguish images with fog present from those free of fog. These image descriptors are global and describe the entire image using Gabor filters at different frequencies, scales and orientations. Our experiments demonstrated hight potential of the proposed method for fog detection on daytime images.

Fast Fog Detection for De-Fogging of Road Driving Images

“…Alami et al used the correlation between saturation and RGB color channel in the HSV color model [16]. Pavlic et al trained features based on the power spectrum of the image through SVM [8]. The above-mentioned fog detection algorithms utilize pretraining information as well as self-information of input images, and have limitations that sometimes require specific hardware.…”

“…In order to mitigate or solve the problems, fog detection may be considered as a pre-processing prior to fog removal step. Recently, various fog detection algorithms have been proposed [8]- [16]. Conventional fog detection algorithms require camera environment information or require prior learning.…”

Air Pollution and Fog Detection through Vehicular Sensors

“…These approaches are not reliable for everyday use. A new method is to use the only reliable visible attribute of foggy weather conditions : the decrease in contrast and blurring in the whole image [1]. The power spectrum being the squared magnitude of the Fourier transform for the image that holds information about the frequencies in the image.…”

“…[1] In the preprocessing step the image will be filtered and normalized to reduce illu mination effects and to avoid that some image regions dominate the spectrum. For the normalizat ion a square image section is used and the intensity values are equalized.…”

“…Before the Fourier transformation a windowing with a 2 d imensional Hanning window is applied [1]. In the last step of the Classification Process, a SVM (Support Vector Machine) with a Radial Basis Function (RBF) kernel is used.…”