DehazeNet is an end-to-end system for single image haze removal. It uses a deep convolutional neural network (CNN) to estimate the medium transmission map from a hazy image, which is then used to recover a haze-free image via atmospheric scattering model. The key components of DehazeNet include a Maxout unit for feature extraction and a novel nonlinear activation function called BReLU, which improves the quality of the haze-free image. The system is trained on synthesized data based on the atmospheric scattering model and achieves superior performance compared to existing methods. DehazeNet is efficient and easy to use, with a lightweight architecture that allows for real-time processing without the need for GPUs. The system is tested on benchmark images and synthetic data, showing excellent results in terms of transmission estimation, image restoration, and robustness to varying conditions such as haze density, atmospheric light, and image scale. DehazeNet is also effective in real-world images, particularly in challenging scenarios involving white or gray regions that are difficult to handle with existing methods. The system's performance is evaluated using various metrics, including mean squared error (MSE), structural similarity (SSIM), peak signal-to-noise ratio (PSNR), and weighted peak signal-to-noise ratio (WPSNR). DehazeNet outperforms existing methods in most of these metrics, demonstrating its effectiveness in haze removal. The system is also robust to variations in scattering coefficient, atmospheric light, and image scale, making it a versatile solution for single image haze removal.DehazeNet is an end-to-end system for single image haze removal. It uses a deep convolutional neural network (CNN) to estimate the medium transmission map from a hazy image, which is then used to recover a haze-free image via atmospheric scattering model. The key components of DehazeNet include a Maxout unit for feature extraction and a novel nonlinear activation function called BReLU, which improves the quality of the haze-free image. The system is trained on synthesized data based on the atmospheric scattering model and achieves superior performance compared to existing methods. DehazeNet is efficient and easy to use, with a lightweight architecture that allows for real-time processing without the need for GPUs. The system is tested on benchmark images and synthetic data, showing excellent results in terms of transmission estimation, image restoration, and robustness to varying conditions such as haze density, atmospheric light, and image scale. DehazeNet is also effective in real-world images, particularly in challenging scenarios involving white or gray regions that are difficult to handle with existing methods. The system's performance is evaluated using various metrics, including mean squared error (MSE), structural similarity (SSIM), peak signal-to-noise ratio (PSNR), and weighted peak signal-to-noise ratio (WPSNR). DehazeNet outperforms existing methods in most of these metrics, demonstrating its effectiveness in haze removal. The system is also robust to variations in scattering coefficient, atmospheric light, and image scale, making it a versatile solution for single image haze removal.