论文部分内容阅读
本文介绍了一种稳定而有效的算法，它能够检测出包围某一已知区域的所有凸形。这些凸形的边界可以是不连续的，这使得这种算法在发现由于部分遮挡或噪声等其他原因而使边界中断的凸形物体时更为有效，从而为进一步的识别工作提供了重要的线索。
This article describes a robust and efficient algorithm that detects all convexities that enclose a given area. These convex boundaries may be discontinuous, which makes this algorithm more efficient in detecting convex objects that interrupt the boundary due to other obstructions such as partial occlusion or noise, providing important clues for further identification .