Image identification method and imaging apparatus
An image identification method for classifying block images of input image data into one of predetermined categories; the method includes the steps of: dividing image data into multiple blocks to produce block images, processing the feature quantity of each block image by their color space information and frequency component, learning separating hyperplanes that indicate boundaries of each category by reading in training data image that have labeled categories for each block and processing image feature quantity for each block of an training data image, and classifying respective block image to a category according to the distance from the separating hyperplane of each category for a newly acquired image to obtain the image feature quantity of block images. An imaging apparatus implementing the image identification method noted above is also disclosed.
1. An image identification method, comprising steps of:
learning separating hyperplanes respectively indicating boundaries of each category by reading a training data image including a plurality of blocks of the training data image to each of which a category is labeled and calculating image feature quantity of each of the plurality of blocks of the training data image;
dividing a newly acquired image into a plurality of blocks to produce a plurality of block images;
calculating the image feature quantity of each of the plurality of block images of the newly acquired image by its color space information and frequency component;
determining a category for each of the plurality of block images of the newly acquired image based on a distance of the image feature quantity from the separating hyperplane of each category;
grouping neighboring blocks of the same category together to form a group of blocks; and
checking whether the determined category for each of the groups of blocks meets inconsistent condition for judging inconsistency of physical relationship between categories stored in an inconsistent condition table or not, based on physical relationship between a targeted group of blocks and other group of blocks surrounding the targeted group of blocks and for which other category is determined.
2. The image identification method according to claim 1 , further comprising a step of:
outputting a result of the step of checking whether the determined category for each of the groups of blocks meets the inconsistent condition or not as a message.
3. The image identification method according to claim 1 or 2 , further comprising steps of:
checking whether a predetermined number or more of surrounding blocks belong to a different category from that of each of the blocks or not, and
changing the determined category of each of the blocks to the different category when the predetermined number or more of the surrounding blocks belong to the different category.
4. The image identification method according to claim 1 or 2 , further comprising a step of:
determining the determined category for each of the groups of blocks to be invalid when the number of blocks included in each of the groups is below or equal to a predetermined number previously set for the category.
5. An imaging apparatus which identifies image category by implementing the image identification method according to claim 1 or 2 , comprising:
an imaging condition storage means for memorizing the parameters for each category of the imaging apparatus and image processing methods after capturing the image;
an imaging means for setting parameters of the imaging apparatus according to the category of subject output by the image identification method when capturing an image, and acquiring a captured image;
and an image processing means for executing image processing by image processing methods corresponding to the category of the captured image.
6. An imaging apparatus which identifies image category by implementing the image identification method according to claim 1 or 2 , comprising:
a program storage means for storing programs that are to be executed with relation to a category;
a program selection means for selecting a program related to the category identified by the image identification method if the program storage means is storing the program; and a program execution means for executing the program selected by the program selection means.