Abstract: The present invention relates to a method and device for obtaining an image of a crumpled document from an image of this document when it is crumpled. The method comprises a step of determining a three-dimensional geometric model of the surface of the crumpled document by triangulation of three-dimensional points defined from the pattern of a target extracted from the image of this document when it is crumpled; the method is characterized in that it comprises a step of determining a projection of the three-dimensional geometric model onto a so-called acquisition plane, by means of error minimization of this projection under constraints of preserving defined geometric characteristics in the vicinity of the three-dimensional points, and in that it comprises a step of superimposing the textures associated with the three-dimensional model onto the projection of this model thus determined.

Abstract: The invention relates to a document-processing system, which includes a device suitable for supplying consecutive recorded images as well as a background surface, comprising a step of storing a standard background surface image. The next step consists of detecting the presence of a document moving across a series of recorded images relative to the standard background surface image. The following step consists of monitoring the movement of the document within the series of images. The end of the movement of the document is then detected and the standard background surface image is updated. The movement of the document is monitored by determining document-extraction masks respectively for the recorded images, the extraction mask of a recorded image being obtained according to an entropy card of the recorded image relative to said standard background surface image.

Abstract: The present invention relates to a method and device for obtaining an image of a crumpled document from an image of this document when it is crumpled. The method comprises a step of determining a three-dimensional geometric model of the surface of the crumpled document by triangulation of three-dimensional points defined from the pattern of a target extracted from the image of this document when it is crumpled; the method is characterised in that it comprises a step of determining a projection of the three-dimensional geometric model onto a so-called acquisition plane, by means of error minimisation of this projection under constraints of preserving defined geometric characteristics in the vicinity of the three-dimensional points, and in that it comprises a step of superimposing the textures associated with the three-dimensional model onto the projection of this model thus determined.

Abstract: The invention relates to a document-processing system, which includes a device suitable for supplying consecutive recorded images as well as a background surface, comprising a step of storing a standard background surface image. The next step consists of detecting the presence of a document moving across a series of recorded images relative to the standard background surface image. The following step consists of monitoring the movement of the document within the series of images. The end of the movement of the document is then detected and the standard background surface image is updated. The movement of the document is monitored by determining document-extraction masks respectively for the recorded images, the extraction mask of a recorded image being obtained according to an entropy card of the recorded image relative to said standard background surface image.

Abstract: A system and method for generating synthetic speech, which operates in a computer implemented Text-To-Speech system. The system comprises at least a speaker database that has been previously created from user recordings, a Front-End system to receive an input text and a Text-To-Speech engine. The Front-End system generates multiple phonetic transcriptions for each word of the input text, and the TTS engine uses a cost function to select which phonetic transcription is the more appropriate for searching the speech segments within the speaker database to be concatenated and synthesized.

Abstract: A system and method for generating synthetic speech, which operates in a computer implemented Text-To-Speech system. The system comprises at least a speaker database that has been previously created from user recordings, a Front-End system to receive an input text and a Text-To-Speech engine. The Front-End system generates multiple phonetic transcriptions for each word of the input text, and the TTS engine uses a cost function to select which phonetic transcription is the more appropriate for searching the speech segments within the speaker database to be concatenated and synthesized.