Canny detection

Canny detection步骤 1.去噪（Noise reduction）：用高斯滤波器减少图像中的噪声 2.梯度计算（Gradient computation）：计算每个像素的梯度大小和方向（可以用sobel算子） 3.非极大值抑制（Non-maximum suppression）：抑制不属于边缘的像素（将每个像素的大小与其八领域内的像素大小进行对比，如果是局部最大值，就保留，反之置为0） 4.双重阈值处理（Double thresholding）：设置两个阈值，高阈值和低阈值。如果像素的梯度大小高于高阈值，则认为是强边缘，如果像素梯度大小小于低阈值，则置为0。如果像素梯度大小介于二者中间，则认为是弱边缘 5.连接边缘（Edge tracking by hysteresis）：遍历每一个强边缘，将其八领域内的弱边缘与之连接，形成一组连续的曲线

Canny边缘检测被认为是最优的，因为它被设计为最小化检测到的假边缘的数量，同时最大化检测到的真边缘的数量。它通过使用高斯平滑、梯度计算、非最大抑制和迟滞阈值的组合来实现这一点。这些步骤共同作用于抑制噪声和像素强度的微小变化，同时保留图像中的强边缘。

2020-J-2c Outline the steps involved in the Canny edge detection process. In what sense in Canny edge detection optimal	2019-J-3c （类似）List and explain the four steps involved in the Canny edge detector. In terms of edge detection, what trade-off does the detector try to optimize?
1. 用高斯的导数过滤图像 2. 找到梯度的大小和方向 3. 非最大限度的压制。 - Thin wide “ridges” down to single pixel width 4. 连接和阈值处理（hysteresis）。 - 定义两个阈值：低和高 使用高阈值来找出强边缘线，使用低阈值来延续它们。（把有连接关系的边找回来）	1. Filter image with derivative of Gaussian 2. Find magnitude and orientation of gradient 3. Non-maximum suppression: • Thin wide “ridges” down to single pixel width 4. Linking and thresholding (hysteresis): • Define two thresholds: low and high • Use the high threshold to start edge curves and the low threshold to continue them
Canny edge detection is considered to be optimal because it is designed to minimize the number of false edges detected while simultaneously maximizing the number of true edges detected. These steps work together to suppress noise and small variations in pixel intensity, while simultaneously preserving the strong edges in the image.	Canny边缘检测被认为是最优的，因为它被设计为最小化检测到的假边缘的数量，同时最大化检测到的真边缘的数量。它通过使用高斯平滑、梯度计算、非最大抑制和迟滞阈值的组合来实现这一点。这些步骤共同作用于抑制噪声和像素强度的微小变化，同时保留图像中的强边缘。
The Canny edge detection process consists of the following steps: Noise reduction: Reduce noise in the image using a Gaussian filter. Gradient computation: The gradient of the image is computed using a Sobel operator or similar method. This step calculates the magnitude and direction of the gradient at each pixel, which provides information about the edges in the image. Non-maximum suppression: After the gradient has been computed, the algorithm applies non-maximum suppression to suppress any pixels that are not part of an edge. This step involves comparing the gradient magnitude at each pixel to the gradient magnitudes of its neighbors and setting the pixel to zero if it is not a local maximum. Double thresholding: In this step, the algorithm applies two thresholds to the gradient magnitude image. Pixels with a gradient magnitude above the higher threshold are considered to be strong edges, while pixels with a gradient magnitude between the two thresholds are considered to be weak edges. Pixels with a gradient magnitude below the lower threshold are set to zero. Edge tracking by hysteresis: Finally, the algorithm tracks the strong edges and connects them to the weak edges that are connected to them, resulting in a set of continuous edge curves. This step is known as hysteresis thresholding.	Canny边缘检测过程包括以下步骤。 降噪。 使用高斯滤波器减少图像中的噪声。 梯度计算。使用Sobel算子或类似方法计算图像的梯度。这一步计算每个像素的梯度的大小和方向，从而提供关于图像中的边缘的信息。 非最大限度的压制。在计算出梯度后，该算法应用非最大抑制法来抑制不属于边缘的任何像素。这一步包括将每个像素的梯度大小与它的邻居的梯度大小进行比较，如果该像素不是局部最大值，则将其设置为零。 双重阈值处理。在这一步，算法对梯度图像应用两个阈值。梯度大小高于高阈值的像素被认为是强边缘，而梯度大小在两个阈值之间的像素被认为是弱边缘。梯度大小低于低阈值的像素被设置为零。 通过滞后进行边缘跟踪。最后，该算法追踪强边缘，并将它们与与之相连的弱边缘连接起来，从而形成一组连续的边缘曲线。这个步骤被称为滞后阈值处理。