A Content Based Image Retrieval using Soft Computing Technique

Traditional ways of image retrieval using the keywords has many problems associated with that. The flourishing concern in the area of CBIR has been entrenched in latter years. The CBIR system follows the approach to retrieve images based on their visual content. CBIR system achievement primarily depends on specific image depiction and identical operations. A new technique i.e. soft computing is used in this system which will bring accurate results as to the previous proposed system. To get the desired range of similarity, this technique finds the similarity of data store in each image to query image in terms of its visual features and gives the output image. To build an efficient NN feature, we need the functionality to auto calculate the weights according to the data sets. To achieve this, we need to apply back propagation or feed forward algorithm for neural network classification and further, for feature matching, we need to calculate cross correlation and apply regression Model.