PROJECT TITLE :

HazDesNet An End-to-End Network for Haze Density Prediction

ABSTRACT:

Weather conditions ought to be taken into consideration by vision-based intelligent systems like those used for driver assistance and transportation. The appearance of haze in photographs can represent a significant safety risk in driving situations. The visibility and usability of hazy images captured in real-world conditions is measured using something called haze density. The forecasting of haze density can be useful for a variety of vision-based intelligent systems, particularly those systems that are used in environments that are exposed to the outdoors. Predicting the density of the haze is a difficult task because the haze and many other components of the scene have a lot in common in terms of their appearance. Existing methods typically make use of a variety of priors and the design of intricate, manually crafted features in order to make predictions regarding the visibility or haze density of an image. In this article, we propose a novel method to predict haze density that is based on an end-to-end convolutional neural network (CNN) and is given the name HazDesNet. As input, our HazDesNet algorithm receives a hazy image and outputs a pixel-level haze density map prediction. After this, the density map is refined and smoothed, and the average of the refined map is used to determine the overall haze density of the image. A subjective human study is performed to build a Human Perceptual Haze Density (HPHD) database, which includes 500 real-world hazy images and 100 synthetic hazy images, as well as the corresponding human-rated perceptual haze density scores for each set of images. This is done so that the performance of HazDesNet can be verified. The results of our experiments indicate that our method achieves the best performance for haze density prediction on both our newly built HPHD database and other databases already in existence. In addition to providing quantitative results for the entire planet, our HazDesNet system is also able to forecast haze density maps that are continuous, stable, fine, and high-resolution.


Did you like this research project?

To get this research project Guidelines, Training and Code... Click Here


PROJECT TITLE : Traffic Signal Control Using End-to-End Off-Policy Deep Reinforcement Learning ABSTRACT: However, road intersections have historically been among the most significant traffic bottlenecks that have contributed
PROJECT TITLE : Towards End-to-End Text Spotting in Natural Scenes ABSTRACT: The ability to recognize text within images of natural scenes is critical for a wide variety of image understanding tasks. Text detection and recognition
PROJECT TITLE : End-to-End Automatic Morphological Classification of Intracranial Pressure Pulse Waveforms Using Deep Learning ABSTRACT: Objective. The mean intracranial pressure, also known as ICP, is a measurement that is frequently
PROJECT TITLE : End-to-End Automatic Morphological Classification of Intracranial Pressure Pulse Waveforms Using Deep Learning ABSTRACT: Objective. The mean intracranial pressure, also known as ICP, is a measurement that is
PROJECT TITLE : An End-to-End Multi-Task Learning Model for Drivable Road Detection via Edge Refinement and Geometric Deformation ABSTRACT: In this paper, an end-to-end neural network model is used to present a road detection

Ready to Complete Your Academic MTech Project Work In Affordable Price ?

Project Enquiry