The invention discloses a water surface object salvage method and salvage ship based on OpenCV and deep learning, and the method comprises the steps: obtaining an image of a region in front of a ship body, carrying out the preprocessing of the obtained region image through employing an OpenCV image preprocessing method, and obtaining a region image with an object rectangular contour; secondly, target detection is conducted on the image obtained in the first step through a trained Mobilenet SSD model, only an object to be fished is operated, and a ship body is driven to be close to an object in front of the ship body till the real-time distance L between the ship body and the object in front of the ship body is within the range where a mechanical arm can stretch to touch; a current ship body front area image is obtained, whether the center of a current object is located at the center of the area image or not is judged, and if yes, a mechanical arm is controlled to conduct grabbing action; and if not, adjusting the position of the ship body according to the position relationship between the current object center and the regional image center.
本发明公开了一种基于OpenCV与深度学习的水面物体打捞方法及打捞船,包括:获取船体前方区域图像,采用OpenCV图像预处理方法对得到的区域图像进行预处理,得到具有物体矩形轮廓的区域图像;通过采用训练好的Mobilenet SSD模型对步骤1得到的图像进行目标检测,仅对待打捞物体进行操作,驱动船体靠近船体前方物体,直至船体与船体前方物体之间的实时距离L在机械臂可伸长至触碰的范围;获取当前船体前方区域图像,判断当前物体中心是否位于区域图像中心处,若处于,则控制机械臂进行抓取动作;若不处于,则根据当前物体中心与区域图像中心的位置关系,调整船体位置。
Water surface object salvage method and salvage ship based on OpenCV and deep learning
一种基于OpenCV与深度学习的水面物体打捞方法及打捞船
2022-03-25
Patent
Electronic Resource
Chinese
IPC: | G06V / G06N COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS , Rechnersysteme, basierend auf spezifischen Rechenmodellen / B63B Schiffe oder sonstige Wasserfahrzeuge , SHIPS OR OTHER WATERBORNE VESSELS / G06K Erkennen von Daten , RECOGNITION OF DATA / B63C LAUNCHING, HAULING-OUT, OR DRY-DOCKING OF VESSELS , Zuwasserlassen, Trockenstellen, Docken oder Bergen von Schiffen |
Engineering Index Backfile | 1943
|