You can enable content indexing and entity detection for images in additional file types. By default, only scanned documents in PDF format are processed by using optical character recognition (OCR).
To the content analyzer computer, add the OCRFilterFileList additional setting as shown in the following table.
A comma-separated list of file extensions that contain images. For example, you can enter the following comma-separated list of file extensions: pdf,jpeg,docx,doc,msg.