User profiling is a task to analyze the mentality of a person from his/ her different activities. In this era of social media, people perform different activities on different social media platforms. By analyzing these activities, the sentiments of the corresponding users are retrieved which are used in different target specific purposes. On social media platforms, people generally post three types of contents: Textual messages, Images, or Videos. In this proposed work, sentiment analysis has been performed based on the images posted by a user on the social media platform. This whole work comprised of two phases. In the phase-1 the captions of the images have been retrieved using the deep learning approach, and in phase-2 the sentiment of the user has been retrieved by analyzing the contexts of the captions. For the experiment, the X handles or the Facebook handles of a few users have been analyzed. The results and the challenges are explained in detail in this report.