Understanding position parameters to embed forensic watermarks in video frames helps protect premium content on OTT platforms

Embedding watermarking enhances the security of DRM-protected premium video content hosted on the cloud. There are a set of parameters that specify the position and size of the watermark on the output media object. They help in dealing with the practical challenges of selecting the most suitable watermark position, along with maintaining robustness and efficiency in the process.

Each parameter has specific usage in specifying the position and size of the watermark in the content stream and differs for image-based and text-based video watermarking schemes. These parameters are discussed below.

ReferPos is the approximate position of the watermark relative to the output stream. It can have four values based on watermark position: Upper left, lower left, upper right, and lower right.

Dx represents the horizontal shift of the watermark vis-a-vis the output video image. Dx can either be absolute – a positive integer (in pixels, ranging from 8 to 4096), or a proportion value (between 0 to 1). Dy value represents the vertical shift of the watermark on the video image. It can have absolute value or proportion like Dx. A value of 0.5 represents a 50% horizontal (in case of Dx) and/or vertical (in case of Dy) shift of watermark in proportion to DRM protected content width and/or height. Default values of both Dx and Dy are 0. The Dx and Dy values should always be lesser than the size of the output video image, else the watermark will display partially.

Likewise, width and height parameters depict the width and height of the watermark on the output video stream, which can again be absolute or relative. In case not set, the watermark width is 0.12 times the output video width. In case only one of the values is present, the other parameter is calculated based on the aspect ratio of the source watermark image. In case the resolution of the output video stream changes frequently, the position and size should be set to the video’s proportion values.

Timeline parameters determine the start time and duration of a watermark on the video stream. Both are mandatory fields, string type, and expressed in seconds. The default value of start is 0, and duration is default from the start till the end of the video.

In the case of text watermarks, there are only two parameters: The top value represents the watermark position offset with respect to the top position, and the left value represents offset from the left position. Both can accept only positive integer values.

The post Understanding position parameters to embed forensic watermarks in video frames helps protect premium content on OTT platforms first appeared on Giga Biting.