Search

The purpose of this work is to develop a robust technique for image inpainting to restore small cracks as well as large regions those include the regions developed by object removal. The image to be restored is transformed into sub-bands by the use of DWT (Discrete Wavelet Transform). These sub-bands are reconstructed back to spatial domain to obtain the subspaces images that are at the same scale as the original image to be restored but having different resolutions. These subspace images are then subjected individually to ‘prioritized exemplar approach’ to fill-in different structures and textures simultaneously. We also optimize the patch size to cope up different sizes of textures, structures and varying resolution of the subspace images. These restored subspace images are superposed to obtain the final restored image. A number of images with changing complexion are used to estimate the effectiveness of the algorithm. The results shows visually plausible background where from the object is removed in variety of images with different structures and textures. The RMSE (Root Mean Squared Error) and PSNR (Peak Signal to Noise Ratio) measures are used to quantify the improvement over visual quality of the restoration.

This paper presents a novel feature extraction framework for content-based image retrieval (CBIR). Discrete wavelet transform (DWT) based Local tetra pattern (LTrP) is used to obtain the feature map from an input image. Decomposition of DWT up to single level and the features obtained from it would make the CBIR system very sensitive to noise. Therefore, decomposition up to three scales is used to remove noise. On each of the sub band LTrP is applied and 130 features are extracted. Further, Artificial Neural Network (ANN) is employed for index matching and image retrieval task which gives the classification accuracy of 97.9 % for Corel 1K database. We have compared our proposed feature extraction scheme with LTrP and other existing local descriptor. Results show that combination of DWT and Local Tetra Patterns (DWT + LTrP) extracts more robust features than LTrP alone. Also, the effect of different wavelet filters on the accuracy of the system has been analyzed. Proposed framework has been tested on Corel-1000 and Corel-10000 databases. We have used average retrieval rate, precision, and recall for performance measure of the CBIR system. Proposed method outperforms the retrieval rate from 75.9% using LTrP to 97.9% using proposed method on Corel 1K database. Improvement in performance measures as compared to the other existing methods evidence that, the proposed spatio-frequency local descriptor is more robust for image retrieval task.

Recovering dehazed image from the single underwater image has always been a challenging task. Inspired from dark channel prior, we extend it further to dehaze the underwater images using a single color channel. The red color component undergoes maximum attenuation on account of the longest wavelength and dominates the dark channel prior in the underwater scenario. We developed the algorithm based on the red color channel, and obtained depth map using morphological operation. The proposed approach significantly restores the color, minimizes the effect of haze and improves the contrast. Experimental results based on qualitative and quantitative analysis exhibits that the proposed method performs efficiently as compared to existing dehazing algorithms.

The use of computer vision based approach has started to bring the intelligence to many of the modern machineries. Such kind of high performance image processing systems can be efficiently built using Field Programmable Gate Arrays (FPGAs). This paper presents the design and implementation of FPGA based Computer Vision System for offset error computation of a new proposed registration mark pattern in 4-color web offset printing machines. The color printing quality of offset machine degrades due to a genuine problem of registration error caused by improper alignment of the prints from each process color section. This system can be used in an automated registration control system for web printing press that will control the position of each of CMYK cylinders depending on offset error calculated which will improve printing quality.