The Image Plugin is an Content Based Image Retrieval Plugin for Elasticsearch using LIRE (Lucene Image Retrieval). It allows users to index images and search for similar images.

It adds an image field type and an image query

See http://demo.elasticsearch-image.com for a demo of the plugin

In order to install the plugin, simply run: bin/plugin install https://github.com/shi-yuan/elasticsearch-image/releases/download/v2.3.1/elasticsearch-image-2.3.1.0.zip.

curl -XPUT 'localhost:9200/test/test/_mapping' -d '{
    "test": {
        "properties": {
            "my_img": {
                "type": "image",
                "feature": {
                    "CEDD": {
                        "hash": "BIT_SAMPLING"
                    },
                    "JCD": {
                        "hash": ["BIT_SAMPLING", "LSH"]
                    },
                    "FCTH": {}
                },
                "metadata": {
                    "jpeg.image_width": {
                        "type": "string",
                        "store": "yes"
                    },
                    "jpeg.image_height": {
                        "type": "string",
                        "store": "yes"
                    }
                }
            }
        }
    }
}'

type should be image. Mandatory

feature is a map of features for index. Mandatory, at least one is required

hash can be set if you want to search on hash. Optional

metadata is a map of metadata for index, only those metadata will be indexed. See Metadata. Optional

curl -XPOST 'localhost:9200/test/test' -d '{
    "my_img": "... base64 encoded image ..."
}'

curl -XPOST 'localhost:9200/test/test/_search' -d '{
    "query": {
        "image": {
            "my_img": {
                "feature": "CEDD",
                "image": "... base64 encoded image to search ...",
                "hash": "BIT_SAMPLING",
                "boost": 2.1,
                "limit": 100
            }
        }
    }
}'

feature should be one of the features in the mapping. Mandatory

image base64 of image to search. Optional if search using existing image

hash should be same to the hash set in mapping. Optional

limit limit the number of results returned (per shard) for scoring. Optional, only works when hash is specified

boost score boost Optional

curl -XPOST 'localhost:9200/test/test/_search' -d '{
    "query": {
        "image": {
            "my_img": {
                "feature": "CEDD",
                "index": "test",
                "type": "test",
                "id": "image1",
                "path": "my_image",
                "hash": "BIT_SAMPLING"
            }
        }
    }
}'

index the index to fetch image from. Default to current index. Optional

type the type to fetch image from. Mandatory

id the id of the document to fetch image from. Mandatory

path the field specified as path to fetch image from. Mandatory

routing a custom routing value to be used when retrieving the external image doc. Optional

byte[] bytes;
try (BytesStreamOutput os = new BytesStreamOutput()) {
    ImageIO.write(ImageIO.read(file), formatName, os);
    bytes = os.bytes().toBytes();
}

ImageQueryBuilder query = new ImageQueryBuilder("img");  //image field
query.image(bytes);
query.feature(feature);
query.hash(hash);
query.boost(boost);
query.limit(limit);
query.lookupIndex(INDEX);
query.lookupType(TYPE);
query.lookupId(itemId);	
query.lookupPath(path);
query.lookupRouting(routing);

SearchRequestBuilder queryBuilder = searchClient.prepareSearch(INDEX).setTypes(TYPE)
    .setSearchType(SearchType.DFS_QUERY_THEN_FETCH)
    .setQuery(query)
    .setFrom(from)
    .setSize(size);

Metadata are extracted using metadata-extractor. See SampleOutput for some examples of metadata.

The field name in index will be directory.tag_name, all lower case and space becomes underscore(_). e.g. if the Directory is JPEG and Tag Name is Image Height, the field name will be jpeg.image_height

Images are processed by Java ImageIO, supported formats can be found here

Additional formats can be supported by ImageIO plugins, for example TwelveMonkeys

AUTO_COLOR_CORRELOGRAM, BINARY_PATTERNS_PYRAMID, CEDD, SIMPLE_COLOR_HISTOGRAM, COLOR_LAYOUT, EDGE_HISTOGRAM, FCTH, GABOR, JCD, JOINT_HISTOGRAM, JPEG_COEFFICIENT_HISTOGRAM, LOCAL_BINARY_PATTERNS, LUMINANCE_LAYOUT, OPPONENT_HISTOGRAM, PHOG, ROTATION_INVARIANT_LOCAL_BINARY_PATTERNS, SCALABLE_COLOR, TAMURA

BIT_SAMPLING, LSH

Hash will increase search speed with large data sets

See Large image data sets with LIRE ?some new numbers

• upgrade to 2.3.1.0

• Use multi-thread when multiple features are required to improve index speed
• Allow index metadata
• Allow query by existing image in index

• Added limit in image query
• Added plugin version in es-plugin.properties

• initial release