Object detection
This disclosure provides a method for object detection. The method comprises receiving a user input that specifies one or more first regions and one or more second regions in a template image. The one or more second regions include one or more objects of interest. The method further comprises for each of the one or more first regions discovering a third region in an image under detection corresponding to the first region in the template image by matching the image under detection with the template image. The method further comprises computing a transformation function based on the matching from each of the one or more first regions to its corresponding third region. The method further comprises applying the computed transformation function to the one or more second regions to localize one or more fourth regions in the image under detection for the object detection.
1. A processor-implemented method for object detection, the method comprising:
capturing an image under detection, wherein the image under detection is of a product;
receiving a user input that specifies one or more first regions and one or more second regions in a template image, wherein the one or more first regions comprise one or more salient parts for matching, wherein the salient parts comprise matrix barcodes or invariant patterns, and wherein the one or more second regions include one or more objects of interest;
for each of the one or more first regions, finding a third region in an image under detection corresponding to the first region in the template image by matching the image under detection with the template image by performing template matching via search, wherein template matching via search comprises comparing the image under detection against the template image;
computing a transformation function based on the matching from each of the one or more first regions to its corresponding third region, wherein the transformation is either a two-dimensional or three-dimensional transformation, and wherein a subset of the user-specified first regions are selected for use in computing the transformation function based on a level of similarity between a user-specified first region and the corresponding third region in the image under detection;
applying the computed transformation function to the one or more second regions to localize one or more fourth regions in the image under detection for the object detection; and
detecting a defect in the image under detection, based on the one or more localized fourth regions.