CBIR-ACHS: compressed domain content-based image retrieval through auto-correloblock in HEVC standard

Because the complete decompression of images encoded through the HEVC standard is time-consuming, image retrieval encoded through the HEVC standard is one of the most challenging topics in existing database management. Our proposed method is a content-based image retrieval method in the compressed domain, leveraging feature vectors constructed from the I-frame information encoded in HEVC. We introduce a novel concept called auto-correloblock to form these feature vectors. These feature vectors, which represent HEVC coded images in the compressed domain, are based on block sizes and intra-prediction modes. In our method, we utilize the feature vector of the query image to generate a histogram, which is then compared to the feature vectors of images stored in the database. To evaluate our approach, we conduct experiments using the INRIA HOLIDAY database. The results demonstrate the effectiveness of our proposed method, achieving an ANMRR (Average Normalized Modified Retrieval Rank) of 0.23 in coded image retrieval and outperforms its counterparts in this domain.