Co-word mapping of Image Retrieval based on Web of Science-Indexed Papers

Background and aim: Given the special status and wide usage of image retrieval in various fields, the present investigation studied on the research trends and significant factors within the field of image retrieval and drawing the co-word map based on the articles indexed in Web of Science. Material and methods: This scientometric study was performed using bibliometric techniques such as co-citation analysis. Samples of the current study were all articles indexed in ISI in the field of image retrieval from 2001 to 2012. Therefore, 2537 papers were retrieved in this field. Citespace and VOSviewer were applied for co-word analysis. Findings: The highest centrality with the number of 0.18 has been related to the term "image retrieval". "Content based image retrieval" and "relevance feedback" both with 0.15 centrality have been in the next rank. The Highest burst with the number of 11.59 was belonged to "pattern recognition society". "Content-based image retrieval", "image database" with the number of 7.53 and 5.79 burst have won the second and third ranks, respectively. Sigma was obtained 1.39 for the "shape" and 1 for other terms during this period. Also, the analysis of co-word network in VOSviewer indicated 9 scientific clusters in the field of image retrieval. Conclusion: The analysis of co-word network in the field of image retrieval have shown that the content-based image retrieval is one of the most important approaches in the field of image retrieval in the past years.