无人驾驶（二）| lane detection | Udacity | Gradient到底是个啥？（通过线性代数矩阵进行分析）

Gradient曾经是个让我迷惑的词，现在通过Matrix来看看

先来看看opencv的代码

gray = cv2.cvtColor(im, cv2.COLOR_RGB2GRAY)
sobelx = cv2.Sobel(gray, cv2.CV_64F, 1, 0)
sobely = cv2.Sobel(gray, cv2.CV_64F, 0, 1)

可以看到Sobel函数后面有两个参数0，1。其中（1，0）表示$Sobe{l}_x$，（0，1）表示$Sobe{l}_y$

$gradient=\sum (region\ast S_x)$

同时我们得到了$Gradien{t}_x$和$Gradien{t}_y$

$\mathbf{G}=\sqrt{{{\mathbf{G}}_x}^2+{{\mathbf{G}}_y}^2}$

与此同时我们也可以得到

$\mathbf{\Theta }=\mathrm{atan}\left(\frac{{\mathbf{G}}_y}{{\mathbf{G}}_x}\right)$

接下来我们就可以通过threshold来控制gradient和direction，上代码！

# Apply each of the thresholding functions
# You should write these functions on your own
gradx = abs_sobel_thresh(image, orient='x', sobel_kernel=ksize, thresh=(0, 255))
grady = abs_sobel_thresh(image, orient='y', sobel_kernel=ksize, thresh=(0, 255))
mag_binary = mag_thresh(image, sobel_kernel=ksize, mag_thresh=(0, 255))
dir_binary = dir_threshold(image, sobel_kernel=ksize, thresh=(0, np.pi/2))

combined = np.zeros_like(dir_binary)
combined[((gradx == 1) & (grady == 1)) | ((mag_binary == 1) & (dir_binary == 1))] = 1