Object recognition using morphologically-processed images

A non-transitory computer-readable medium whose contents cause a processing device including a processor to perform a method of recognizing an occurrence of an object in an image, where the appearance of the object may vary from one occurrence to another, the method comprising:

• acquiring a target image;
• processing the target image, by the processor, according to a first set of morphological operations to obtain a set of morphologically-processed target images, the set of morphologically-processed target images including a plurality of morphologically-processed target images;
• defining a model set for object recognition including acts of:
• acquiring a model image;
• processing the model image, by the processor, according to a second set of morphological operations to obtain a set of morphologically-processed model images, the set of morphologically-processed model images including a plurality of morphologically-processed model images;
• and
• creating a model from each of the plurality of morphologically-processed model images, respectively;
• performing object recognition, by the processor, on each of the morphologically-processed target images of the set of morphologically-processed target images for each of the models of the defined model set;
• and
• selecting at least one candidate occurrence of a model of the defined model set in the target image, by the processor, based on the object recognition performed on each of the morphologically-processed target images of the set of morphologically-processed target images.

An apparatus comprising:

• a memory configured to store a model image;
• one or more processors coupled to the memory;
• an image acquisition device coupled to the one or more processors, the image acquisition device configured to acquire a target image;
• and
• a computer storage medium storing instructions that, when executed by the one or more processors, cause the one or more processors to perform operations that recognize an occurrence of an object in the target image, where the appearance may vary from one occurrence to another, the operations including:
• processing the target image according to a first set of morphological operations to obtain a set of morphologically-processed target images, the set of morphologically-processed target images including a plurality of morphologically-processed target images;
• defining a model set for object recognition including acts of:
• acquiring the model image from the memory;
• processing the model image according to a second set of morphological operations to obtain a set of morphologically-processed model images, the set of morphologically-processed model images including a plurality of morphologically-processed model images;
• and
• creating a model from each of the plurality of morphologically-processed model images, respectively;
• performing object recognition on each of the morphologically-processed target images of the set of morphologically-processed target images for each of the models of the defined model set;
• and
• selecting at least one candidate occurrence of a model of the defined model set in the target image based on the object recognition performed on each of the morphologically-processed target images of the set of morphologically-processed target images.

A system comprising:

• an image acquisition device configured to acquire a target image;
• one or more processors coupled to the image acquisition device;
• and
• a computer storage medium storing instructions that, when executed by the one or more processors, cause the one or more processors to perform operations that recognize an occurrence of an object in the target image, where an appearance of the object may vary from one occurrence to another, the operations including:
• processing the target image according to a first set of morphological operations to obtain a set of morphologically-processed target images, the set of morphologically-processed target images including a plurality of morphologically-processed target images;
• defining a model set for object recognition including acts of:
• acquiring a model image;
• processing the model image according to a second set of morphological operations to obtain a set of morphologically-processed model images, the set of morphologically-processed model images including a plurality of morphologically-processed model images;
• and
• creating a model from each of the plurality of morphologically-processed model images, respectively;
• performing object recognition on each of the morphologically-processed target images of the set of morphologically-processed target images for each of the models of the defined model set;
• and
• selecting at least one candidate occurrence of a model of the defined model set in the target image based on the object recognition performed on each of the morphologically-processed target images of the set of morphologically-processed target images.