Carbonate reservoirs are still attracting oil and gas investors given the fact that the largest producing field in Ghawar Field is from carbonate reservoir. However, because of their heterogeneity and complex depositional settings, carbonate body identification poses special challenges. Conventional techniques for detecting carbonate bodies in seismic frequently depend on labor-intensive and subjective manual interpretation which are time consuming and prone to human error. Over the years, the usage of Artificial Intelligence in the workflow of geoscience has attracted the attention of researchers. Image analysis via machine learning or deep learning has promising results for application to exploration and production technologies. The demands for the automation of carbonate body detection in seismic to shorten the delivery time of work have been growing. In this study, the authors propose a hybrid image analysis technique based on deep neural network for carbonate detection in seismic which is an eye-catching process by most of the exploration geophysicists using the Yolov5 model. By employing convolutional neural network (CNNs) algorithm, the proposed method can analyze vast amounts of seismic data with high accuracy and efficiency. In addition, this image analysis is based on pattern detection that combines geological knowledge with machine learning techniques to overcome the challenges in detecting carbonates especially in complex depositional settings. The study concludes that integrating automated detection into seismic interpretation can revolutionize the field, enhancing and reliability of subsurface geological interpretations. This advancement in automation can significantly reduce the time and cost associated with seismic data analysis, providing a more consistent and objective interpretation of carbonate Future research will focus on refining the model by adding segmentation, incorporating more diverse and complex datasets and integration of seismic signals to the detection.