论文部分内容阅读
针对海底集矿机采矿环境图像,采用分段线性变换提高图像细节,中值滤波去除悬浮物干扰.利用形态学抗噪声梯度算子提取地形和障碍物轮廓,并用分段线性拟合计算出地表亮度变化率.根据表面亮度变化特征判断障碍物类型,采用自适应形态学对轮廓进行细化与连接.通过障碍投影变换计算出障碍物的距离、高度和宽度等信息.对陆地图像进行了分析,证明位置、高度和坡度等参数计算的可行性.利用上述方法对深海底的图像进行处理,不仅保留了边界信息,且提高了抗干扰能力和抗边界间相互影响能力,可有效识别深海底地形和障碍物,得出位置和形状等参数,可以为集矿机避障系统信息融合技术提供可靠数据.
Aiming at the mining environment image of submarine collector, the piecewise linear transformation is used to enhance the image details and the median filter is used to remove the interference of suspended matter.The morphological anti-noise gradient operator is used to extract the topography and obstacle profile and the piecewise linear fitting is used to calculate the surface Brightness change rate.According to the change of surface brightness, the type of obstacle is judged, the outline is refined and connected by using adaptive morphology, and the information such as the distance, height and width of obstacle is calculated through the obstacle projection transformation. , Prove the feasibility of the calculation of parameters such as position, height and slope, etc. Using the above method to process the image of the deep seabed not only retains the boundary information, but also improves the anti-jamming ability and the ability to resist the interaction between the boundaries, which can effectively identify the deep sea floor Topography and obstacles, and draws the parameters such as location and shape, which can provide reliable data for information fusion technology of obstacle avoidance system in mine concentrator.