In the field of computer vision, the clarity of the real-time video is affected in hazy conditions where there is a problem in efficiently recognizing the face of the person. This research endeavors to address this challenge mainly through two techniques: image dehazing and face recognition. Haar Cascading Classifier algorithm is used for face detection which is a pre-trained model provided by OpenCV. For image dehazing, a custom image dehazing module is utilized to enhance image clarity by removing atmospheric haze. Further, this module integrates atmospheric scattering model and the Dark Channel Prior to achieve effective dehazing results. To previously built models utilize various deep neural networks to efficiently eliminate haze from any single image. The study tries to enhance this by including this feature into video monitoring systems. Previous approaches have not been applied in video monitoring systems combined with the face recognition in the video. The proposed approach includes the use of parameters such as transmission coefficients and skylight, which contribute to the accuracy. The experimental results upon integrating these two techniques show a significant improvement in the real-time face recognition systems particularly in challenging environmental conditions characterized by haze and low visibility.