Joint Multimodal Aspect Sentiment Analysis (JMASA) aims to jointly extract aspect terms and their associated sentiments from given text-image pairs. Existing methods focus on associating the entire image with the corresponding text. However, redundant information in the image introduces noise, hindering the highlighting of crucial affective visual regions. Additionally, simply utilizing attention mechanisms to adaptively search for the associated sentiment between aspects in a sentence may ignore that sentiment judgments can be easily interfered with by other irrelevant words. To address these challenges, we propose a novel Aspect-Aware Affective Focus Network (AAFN) for multimodal sentiment analysis. Specifically, our model contains an aspect-aware enhancement module that is sensitive to aspect-related semantic information based on syntactic structure and part-of-speech information. Furthermore, we introduce a candidate affective visual focus module that precisely identifies candidate visual sentiment regions under linguistic guidance. To effectively eliminate the semantic gap, we introduce a language-guided fusion module to achieve fine-grained interaction between visual focus and aspect-related information, thereby enhancing the relevance between image-text pairs. Extensive experiments conducted on two benchmark datasets, Twitter-2015 and Twitter-2017, demonstrate the effectiveness of our proposed method.