Naval ship recognition

[摘要] Object recognition is a very interesting task with multiple applications andfor that reason it has been dealt with very intensively in the last years. Inparticular, the application to naval ship pictures may facilitate the work ofthe coastguards or the navy. However, this type of images entails somedifficulties due to their specific environment. Water reflects the light andas a consequence, some areas may presumably show different brightness andcolor. Waves from wind or moving ships pose a problem due to the additionaledges that they produce. The camouflage of ships in the military context isalso an issue to take into account. Therefore, it is difficult to propose asimple method that is valid for every image.

A discussion about which techniques may solve these problems is presented andfinally a combined solution based on contour recognition is suggested. Testimages are preprocessed by histogram stretching. Then, the Canny method isapplied to the image and to the reference contour in order to obtain not onlytheir edges, but also their respective orientations. The problem ofrecognizing the reference contour within the detected edges is addressed bymaking use of the Generalized Hough Transform (GHT).

[学科分类] 电子、光学、磁材料