This paper presents a novel approach for low light object detection leveraging the Retinex theory. In low light conditions, traditional computer vision methods struggle to accurately identify and localize objects due to poor visibility and contrast. The Retinex theory, based on the human visual system, is employed to enhance image brightness and contrast. Our proposed method combines Retinex based pre-processing with a deep learning based object detection framework to address challenges posed by low light environments. The Retinex pre-processing stage adaptively adjusts the illumination and reflectance components of the input image, enhancing object visibility. Subsequently, a state of threat object detection model is employed to identify and locate objects in the preprocessed image. Experimental results on benchmark datasets show that the proposed strategy outperforms existing methods for low light object detection.