Image retrieval using colour and texture features of Regions Of Interest

Vimina, E. R.; Jacob, K. Poulose

doi:10.1109/infrkm.2012.6204981

Cited by 5 publications

(6 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This method has a precision higher than the previous experiments in [37] and [38] of 68.78% and 64.76% respectively for image retrieval in Wang's database for k=10. Table 2 compares the precision results of the images retrieved for k = 10 for experiments done in [37] and [38]. The table also shows the average precision of each class present in Wang's database.…”

Section: Text Similarity Calculationmentioning

confidence: 57%

“…This might be attributed to lack of images belonging to similar category in Flickr8k dataset. The lowest precision value in identifying a class bus is 60% which is higher than the lowest precision of [37] and [38] of 33% and 28% respectively. The precision-recall graph also shows the curve to be above the diagonal of the graph which clearly indicates the model is good.…”

Section: Recallmentioning

confidence: 70%

“…When analyzing the results, the proposed method of image retrieval using image captioning shows higher results compared to the previous experiments. [37] and [38]. Average precision by this method is clearly high compared to the methods in [37] and [38] and the precision of the proposed method remains high even for increased values of k.…”

Section: Text Similarity Calculationmentioning

confidence: 80%

“…As found in all image retrieval experiments, the F1 score of the model decreases as the value of k increases. Image retrieval using image captioning has a higher average precision than other methodologies used in [37] and [38]. The proposed method clearly shows a higher precision in identifying classes such as beach, buildings, elephants, flowers, mountains, and food compared to the other methods.…”

Section: Recallmentioning

confidence: 84%

“…Like A database table is created for the images in the dataset by using color, shape, texture as features and images are retrieved using similarity calculation between the query and images in the dataset by Sreedevi et al [4]. In [5], the above CBIR techniques are used on a region of interest. In the input query, the region of interest is mentioned, and similarity calculation is performed only on the given region of interest in the image to retrieve similar images.…”

Section: Hashing Methodsmentioning

confidence: 99%

See 4 more Smart Citations

Image Retrieval Using Image Captioning

Vijayaraju

View full text Add to dashboard Cite

The rapid growth in the availability of the Internet and smartphones have resulted in the increase in usage of social media in recent years. This increased usage has thereby resulted in the exponential growth of digital images which are available. Therefore, image retrieval systems play a major role in fetching images relevant to the query provided by the users. These systems should also be able to handle the massive growth of data and take advantage of the emerging technologies, like deep learning and image captioning. This report aims at understanding the purpose of image retrieval and various research held in image retrieval in the past. This report will also analyze various gaps in the past research and it will state the role of image captioning in these systems. Additionally, this report proposes a new methodology using image captioning to retrieve images and presents the results of this method, along with comparing the results with past research.

show abstract