Object detection is a critical task in computer vision with many applications. This includes autonomous vehicles, surveillance, and robotics. The current research investigates the evaluation of object detection models, focusing on their accuracy and reliability. Key metrics such as precision, recall, and intersection over union (IoU) are used to measure performance, providing a comprehensive assessment of model capabilities. The study explores various datasets and evaluation techniques designed to mimic real-world scenarios and challenges, such as cluttered scenes, adverse weather conditions, and low-light environments. We assess the effectiveness of various state-of-the-art object detection models and their underlying architectures, examining their strengths and weaknesses across different application domains. The impact of dataset bias, class imbalance, and transfer learning on model accuracy is also considered. We aim to provide insights into optimizing models for specific tasks and environments through custom evaluation metrics and data augmentation strategies. According to the analysis we are achieving about 81.66% of accuracy for person detection.