Automatic Recognition of Query Based Tablet Image using Improved Color Features
A.B. Dhivya, M. Sundaresan · 2019
Now a day's image mining plays a vital role in various technologies in order to find atypical patterns and to extract the useful information. In this paper, the preprocessing and feature extraction steps are mainly focused. The automatic identification, matching and retrieval of tablet image based on the shape and color features present in the tablet images are also discussed. These color features of the tablet images are improved using histogram based techniques like Histogram, Pyramid Histogram and Modified Histogram to match and retrieve the images of the tablet required from the database. The parameters such as accuracy, error rate and execution speed are used to measure the performance of the tablet identification system. MATLAB software is used for the purpose of implementing the results. From the results it is visible that using the modified histogram features, the accuracy, error rate reduction and the speed of execution has been considerably improved.