A Context-Based Searching Technique by Extraction and Fusion of Metadata of Digital Photos
Nafeez Zawad Hossain, Tanjima Nasreen Jenia, Md. Samshad Rahman, Sofen Hoque Anonta, Khandaker Tabin Hasan · 2020
Memories of events hold the essence of human life. To preserve and restore the ailing memories, digital photo is the most popular approach. In this paper, we present a technique to create a semi-automatic photo memorandum for context-based photo visualization and retrieval by extracting, extending and analyzing the metadata contained in digital photos. Today people tend to capture and share any significant memory or information through photographs. This practice of capturing and storing digital photographs is leading the photo collection of an individual grow exponentially. Thus, it would be a strenuous and time consuming work to find a particular photo from this vast collection if the photos are not labeled with their descriptive properties, i.e., the context. Moreover, the value of a photo dissipates if it cannot be retrieved whenever necessary. Since the human mind stores the context of a photo rather than the creation date or its digital information, it would be convenient to fetch photos by utilizing their contextual data. This paper demonstrates an approach of context-based searching for photos, along with metadata generation and diffusion for context building which will preserve and retrieve the memory by labeling them with various data that remains unnoticed in digital photos with the help of internal and external sources. The work is underpinned by a novel search technique to extract photos using natural language.