The following XML shows a single record produced by image hash analysis:
<output> <record> <pageNumber>1</pageNumber> <startTime iso8601="2018-01-05T11:02:49.038357Z">1515150169038357</startTime> <trackname>ImageHash.Result</trackname> <ImageHashResult> <id>5536224b-4fdf-4fee-bb7a-fe47eee9f010</id> <identity> <identifier>MyImage</identifier> <database>ImageHashes</database> <confidence>100</confidence> <metadata> <item> <key>key1</key> <value>value1</value> </item> <item> <key>key2</key> <value>value2</value> </item> </metadata> </identity> </ImageHashResult> </record> </output>
Media Server creates a separate record for each image hash that an ingested image is compared to.
The identity
element of the record describes the image hash (in your training database) that was used for comparison. This element contains the following information:
identifier
element provides the identifier of the image hash.database
element provides the name of the database.confidence
element provides the confidence score for the match, from 0 to 100 where 100 indicates maximum confidence. You can configure Media Server to discard results with a low confidence score by setting the parameter MatchThreshold
(see Identify Duplicate Images).metadata
element contains any custom metadata associated with the image hash. (You can add custom metadata to image hashes in your training database).
|