Inscription Image Retrieval Using Bag of Visual Words

Abstract

This paper presents a technique for efficient and veracious retrieval of ancient inscriptions and manuscripts from a large database of images by using the Bag of Visual Words (BoVW) technique. The proposed method can be used to recognize inscription images across the world. SURF (speeded up robust features) is used as an image feature extractor. A visual vocabulary is created by representing the image as a histogram of visual words which helps in the retrieval process. Usage of SURF ensures scalability, faster processing better results with darkened and blurred images. We demonstrate the method on a combination of 300 inscriptions images comprising of several inscription around the world.

A New Hybrid Algorithm for Telugu Word Retrieval and Recognition

“…The simulated outputs of retrieved word images with a query word image of 'pariksha' has been shown in Fig. 4, we can observe that there is a noise in the input query but still the retrieved word images are more relevant to the input query, which shows the robustness and effectiveness of proposed method where this is not possible by the algorithms presented in the literature [10,24,29,30]. Figs.…”

“…The query images are selected such that (i) They have multiple occurrences in the database, (ii) They are mostly functional words and (iii) They have no stop words. The performances of proposed method are measured by mean Average Precision (mAP) and mean Average Recall (mAR) and the obtained results are compared with the BoVW [10], SIFT+BoVW [30], algorithm presented in [24] i.e., HMM with correlation [24]SURF + BoVW [29] and HCH scheme.…”

“…Recently in [24], the author has presented a correlation and hidden markov model (HMM) based Telugu word spotting with superior performance over conventional word image retrieval systems. In [29], author has presented an algorithm based on speeded up robust features (SURF) with the help of BoVW to improve the performance over BoVW and SIFT + BoVW. However, there is a lack of efficiency, higher complexity, lower precision with large data bases are the limitations of the word image retrieval algorithms presented in the literature.…”

“…To overcome the drawbacks of BoVW [10], SIFT + BoVW [30], algorithm presented in [24], SURF + BoVW [29] and hash coded hamming (HCH) scheme, our present work aims at performing texture feature extraction given in section 3.1, extraction of properties by computing contrast, correlation, energy and homogeneity (discussed in section 3.2), for classification, we utilized iterative partitioned clustering (discussed in section 3.3). We also calculated the statistical parameters such as mean, variance, standard deviation and entropy followed by similarity measurement for further improvement in the proposed retrieval/recognition system.…”

Telugu Word Image Retrieval using Deep learning Convolutional Neural Networks

“…As shown in Figure 5, different kind of Telugu word images like occlusion affected, missing segment, noisy effected, random distortion and missing segment with random distorted images are considered as a query word images. Assessment of proposed TWIR system using DL-CNN is done by computing mean average precision (mAP) and mean average recall (mAR) and compared with the conventional TWIR systems like SIFT-BoVW [14], HMM-C [16], SURF-BoVW [17], GLCM-IPC [18], HWNET v2 [19] and SDM-NSCT [21]. As discussed earlier, simulation analysis is done with several kind of Telugu word images and obtained enhanced mAP and mAR even when the query word images had a kind of unwanted information which might be introduced automatically while acquiring them or manually by a human or even by a printing machine during the scanning procedure.…”