KR20000007020A - Channel sharing using a data hiding method - Google Patents

Abstract

PURPOSE: A method is provided to share transfer and storage channel by inserting a voice signal to an audio data using a data hiding(watermarking) technique. CONSTITUTION: The channel sharing method comprises the steps of: generating a watermark expressing second digital data desiring a sharing; pre-processing in order to insert the watermark into second digital data; inserting the watermark of the second digital data into the first digital data; extracting an original watermark through a reserve processing of the watermark insertion from the first data in which the watermark is inserted(the second data is hid in the first data); and correcting or repairing an error generated in the extracted watermark.

Description

Channel sharing using a data hiding method

The present invention generates a watermark from another second digital video, video or audio data in order to share a transmission or storage channel assigned to transmit the first digital video, video and audio as a parent. It is about how to hide the data.

Related technologies include a technique for reducing the amount of data while maintaining main information from the second digital video, video or audio data, a technique for hiding the reduced data in the first digital video, video or audio data, and a receiving side ( Or a technique for extracting a watermark from the digital video, video or audio data received from the storage device), and a technique for restoring a second digital video, video or audio data from the extracted watermark.

In the conventional method of transmitting and storing digital data, in order to transmit or store different digital data, each channel has to be allocated a unique frequency band or occupy a different memory area, thereby wasting channel and memory. In addition to the need to attach additional data associated with the data (frames) to data at specific times (in the case of video), not only does it require extra channels or memory, but it also synchronizes the two data. Problems also arise.

The present invention aims to solve the above problem by utilizing a watermark technology used for copyright protection of video and audio. The conventional watermarking method relates to a method of inserting a simple binary image such as a random number or a logo (logo) into the data (video or audio) to be copyright-protected so that it can be visually detected or undetectable.

The present invention addresses the problems arising from the realization of channel sharing techniques that have advantages such as saving channel bands or storage memory, ease of synchronization between disparate data, and the feasibility of various additional functions of video or audio transmission / storage. It is for. Specifically, an object of the present invention is a technique for generating a watermark from additional second data, a method of inserting the generated watermark into the first data, and a watermark generated in a watermarking process and a transmission / storage channel. The present invention relates to a method for efficiently correcting and correcting an error of a.

1 is an overall flowchart of a transmitter (or an encoder)

2 is an overall flowchart of a receiver (or a decoder)

In order to achieve the above object, the present invention is composed of three steps. The first step is to insert as many watermarks as possible into the first data, the main data of the transmission and storage medium, to be resistant to errors such as compression, and the second step is from the second data, the sub data of the transmission and storage medium. Creating a watermark. This step effectively compresses the data while maintaining the content of the data, generates an error correction parity for error correction and correction from the compressed data, and formats these compressed data and parity to the frequency characteristics of the first data. ) To generate a watermark. Finally, the third step is to correct and correct the error of the watermark extracted from the received data (fetched from the storage device). In this step, the error correction code attached to the watermark is decoded to correct the corresponding error, and the detected error is to recover the second data from the watermark, and then correct the detected error using the mutual context of the data and the mutual context of the information. .

Hereinafter, the present invention will be described in detail with reference to the accompanying drawings.

Referring to FIG. 1, first digital data to be a parent is received (S101). As a preprocessing step for inserting an invisible watermark in the frequency domain of the first digital data, digital data is frequency-transformed using a discrete cosine transform (DCT) or a discrete fourier transform (DFT) (S102). A region in which the watermark is to be inserted in the frequency domain is determined (S103). On the other hand, the second digital data for sharing the channel is received (S104). The second data can be another digital video, audio, still image, or other auxiliary data. A lossy or lossless data compression technique is applied to generate a watermark from the second digital data (S105). In order to correct various errors that may occur in the compressed second data, parity is generated by applying error correction coding such as RS (Reed-Solomon) (S106). Data formatting is performed using the results of the analysis of the first digital data (e.g., frequency domain or transitioned scene) in order to properly insert the compressed data and parity into the parent first data. (S107). The formatted second digital data is regarded as a watermark and distributed and inserted at an appropriate position of the first digital data (S108). Inverse frequency conversion is performed to restore the coefficients of the frequency domain in which the watermark is inserted into the data of the original temporal or spatial domain (S109). The compression or channel encoding is applied to the first digital data into which the watermark that cannot be detected is inserted and then transmitted or stored (S110) (S111).

Referring to FIG. 2, first data received or acquired from a memory is input (S201). When reverse transmission / channel encoding is performed on this data, first data is concealed (i.e., watermarked) of the second data (S202). The watermark is converted into a frequency domain in order to extract the watermark from the first data (S203). The watermark made of the second data in the frequency domain is extracted (S204). The extracted watermark is damaged by an error generated during transmission encoding (compression) or transmission / storage of the first data. In order to correct this error, error correction encoding is performed after reverse formatting (S207). The second data is reproduced by performing a data compression reverse process from the watermark in which the error is corrected (S208). The second data thus obtained may be different from the original second data S104. In this case, the detected error is corrected using the context with the surrounding data (S209) and the final first data is obtained (S210). On the other hand, the first data from which the second data is removed obtains the first data after inverse transformation (S205) for watermarking (S206).

As described above, in the present invention, when the transmission band is narrow and two different digital data cannot be transmitted or stored at the same time, the channel is shared so that several different data are simultaneously stored or transmitted on one channel. To take advantage of it, reduce the storage area of the storage device that will store the two data by one-half, or attach auxiliary data to the parent's first data to attach auxiliary data without problems of synchronization or loss. Can be.

Claims (4)

In the method of inserting the second data into the first data undetectable, Compressing the second data into a watermark and inserting the watermark into an area in the first data that cannot be detected by the human audiovisual system. The method of claim 1, wherein when generating a watermark from the second data, a parity for error correction encoding is generated and used as part of the watermark to give a robust characteristic to an error that may occur during transmission or storage. How to. The method of claim 1, wherein when the first data is a video, the screen in which the scene transition of the video occurs or the portion or complexity of each screen, etc., is used to take advantage of the human visual masking effect. How to adaptively resize your watermark. A method for restoring the second data inserted into the first data, comprising: correcting an error generated in the extracted watermark by using an error correction technique, or applying an error concealment technique to the detected error How to fix it.