Photo extraction techniques

A method for extracting individual images from a medium contained in a holder having image-holding areas, comprising the steps of:

• (a) scanning the medium and the holder at a relatively low resolution to generate a low-resolution digital representation of the holder and the medium including the individual images thereon;
• (b) processing the low-resolution digital representation by:
• (b)(1) defining borders of the holder, such that all of the image-holding areas and all of the individual images contained therein are within the defined borders;
• (b)(2) applying a smoothen filter to the low-resolution representation;
• (b)(3) detecting edge segments of the image-holding areas;
• (b)(4) detecting and identifying each of the image-holding areas;
• (b)(5) determining the orientation of at least one image-holding area with respect to a reference, and
• (b)(5)(i) if it is determined that the at least one image-holding area is skewed with respect to the reference, correcting the orientation of the at least one image-holding area;
• and
• (b)(6) locating each of the individual images within the image-holding areas;
• and
• (c) generating an index of all individual images identified on the medium.

An apparatus for extracting individual images from a medium contained in a holder having image-holding areas, comprising:

• a scanner for scanning the medium and the holder at a relatively low resolution to generate a low-resolution digital representation of the holder and the medium including the individual images thereon;
• a storage medium in communication with the scanner for storing the low-resolution representation;
• means in communication with the storage medium for processing the low-resolution digital representation, the processing means including:
• means for defining borders of the holder, such that all of the image-holding areas and all of the individual images contained therein are within the defined borders;
• means for applying a smoothen filter to the low-resolution representation;
• means for detecting edge segments of the image-holding areas;
• means for detecting and identifying each of the image-holding areas;
• means for determining the orientation of at least one image-holding area with respect to a reference, and, if it is determined that the at least one image-holding area is skewed with respect to the reference, for correcting the orientation of the at least one image-holding area;
• and
• means for locating each of the individual images within the image-holding areas;
• and
• means for generating an index of all individual images identified on the medium.

A machine-readable medium having a program of instructions for directing a machine to extract images from a medium, the program of instructions comprising instructions for:

• (a) scanning the medium at a relatively low resolution to generate a low-resolution digital representation of the medium and the individual images thereon;
• (b) processing the low-resolution digital representation by:
• (b)(1) defining borders of the medium, such that all of the individual images are within the defined borders;
• (b)(2) applying a smoothen filter to the low-resolution representation;
• (b)(3) detecting edges of each area containing at least one image;
• b)(4) determining, and if necessary correcting, the orientation of the medium;
• and
• (b)(5) locating each of the individual images within its corresponding area in the medium;
• and
• (c) generating an index of all individual images identified on the medium.

A machine-readable medium having a program of instructions for directing a machine to extract individual images from a medium contained in a holder having image-holding areas, comprising instructions for:

• (a) scanning the medium and the holder at a relatively low resolution to generate a low-resolution digital representation of the holder and the medium including the individual images thereon;
• (b) processing the low-resolution digital representation by:
• (b)(1) defining borders of the holder, such that all of the image-holding areas and all of the individual images contained therein are within the defined borders;
• (b)(2) applying a smoothen filter to the low-resolution representation;
• (b)(3) detecting edge segments of the image-holding areas;
• (b)(4) detecting and identifying each of the image-holding areas;
• (b)(5) determining the orientation of at least one image-holding area with respect to a reference, and
• (b)(5)(i) if it is determined that the at least one image-holding area is skewed with respect to the reference, correcting the orientation of the at least one image-holding area;
• and
• (b)(6) locating each of the individual images within the image-holding areas;
• and
• (c) generating an index of all individual images identified on the medium.