Content based Image Retrieval is mostly utilized to extract the pictures from large database. CBIR, which is additionally known as “Query by image” is a technology allowing organizing the computerized Images by their visible attributes. Or, in other words, CBIR is a method for retrieving pictures dependent not on annotations or keywords, but dependent on the feature taken out straightly from the pictures database. CBIR systems are dependent on the utilizations of computer vision methods to the image retrieval issue in huge databases. CBIR is technology of recovering the utmost visually identical pictures to a specified query picture from a cluster or database of pictures. It is useful in a lot of areas like Photography to search images from the database, medical diagnosis etc. Physically annotating the pictures by inputting the metadata or keywords in a huge database can be laborious and might not capture the keyword anticipated to define that picture. CBIR supports in recovering identical pictures from an database of pictures deprived of pictures annotation. In this paper, we are compare the Deep Neural Networks and Neuro-Fuzzy Classifier, both have different outcomes and different results to predict the image. The comparison of our proposed methods Neuro-fuzzy classification and deep neural network shows that the improvement in accuracy. The accuracy values 71.6% and 73.4% for DNN and Neuro-Fuzzy Classifier[Formula: see text] method. The visual and qualitative results are presented for validation of the proposed method.