The recent literature reports several practical and important use cases of social media informatics where artificial intelligence (AI), machine learning (ML), and other relevant technologies are employed to analyze human sufferings and infrastructure damage in natural disasters. While the textual content of social media platforms conveys relevant and useful information during a disaster, social media imagery content has also been proven very effective in analyzing the scale of damage to infrastructures such as roads, bridges, and buildings. Moreover, disaster-related visual content could also be analyzed to extract people’s perceptions, emotions, sentiments, and responses to disasters, which can help different stakeholders, such as humanitarian organizations and policy-makers. Assessing such aspects of disaster events requires effective and efficient image processing methods to process a large amount social media content. This chapter reviews state-of-the-art techniques and shows their utility in processing social media image streams during disaster response for a diversified set of applications. It also highlights the key applications, challenges, available shared resources (datasets and models), tasks, and future research directions. This chapter will provide a ground for future research and a good starting point for the researchers in the domain.