In computer vision, the most active research application is video surveillance. The work done in the area of video surveillance is improved through selecting the best thresholding methods for frame segmentation. The proposed method involves reading video frames, converting each frame to grayscale, and using thresholding techniques to isolate moving vehicles from the background. Thresholding methods are vital for object detection in video surveillance, especially for vehicle monitoring. Adaptive Gaussian thresholding calculates thresholds dynamically for smaller regions of the image. The threshold is computed for each pixel as the weighted mean of the neighboring pixel values adjusted by a constant, handling varying lighting conditions effectively. Otsu’s Thresholding computes a single global threshold by minimizing intra-class variance between the foreground and background, assuming a bimodal histogram. It is simple and computationally efficient. The Black-White Ratio (BWR) and Black-Pixel Ratio (BPR) are used to figure out the content and features of the image and analyze frequency distributions. BWR and BPR provide insights into the distribution of black and white pixels in a given frame or image. BWR and BPR are calculated by converting the image to a binary image using a thresholding technique. The best thresholding method depending upon the average Structural Similarity Index is Otsu Thresholding.