深度图.py

import cv2
import numpy as np


def depth_map(imgL, imgR, sigma=1.3):
    """ Depth map calculation. Works with SGBM and WLS. Need rectified images, returns depth map ( left to right disparity ) """
    # SGBM Parameters -----------------
    # wsize default 3; 5; 7 for SGBM reduced size image; 15 for SGBM full size image (1300px and above); 5 Works nicely
    window_size = 3

    left_matcher = cv2.StereoSGBM_create(
        minDisparity=-1,
        numDisparities=5*16,  # max_disp has to be dividable by 16 f. E. HH 192, 256
        blockSize=window_size,
        P1=8 * 3 * window_size,
        # wsize default 3; 5; 7 for SGBM reduced size image; 15 for SGBM full size image (1300px and above); 5 Works nicely
        P2=32 * 3 * window_size,
        disp12MaxDiff=12,
        uniquenessRatio=10,
        speckleWindowSize=50,
        speckleRange=32,
        preFilterCap=63,
        mode=cv2.STEREO_SGBM_MODE_SGBM_3WAY
    )
    right_matcher = cv2.ximgproc.createRightMatcher(
        left_matcher)  # Block matching disparity map calculation
    # FILTER Parameters
    lmbda = 80000
    visual_multiplier = 6

    wls_filter = cv2.ximgproc.createDisparityWLSFilter(
        matcher_left=left_matcher)
    wls_filter.setLambda(lmbda)

    wls_filter.setSigmaColor(sigma)
    displ = left_matcher.compute(imgL, imgR)  # .astype(np.float32)/16
    dispr = right_matcher.compute(imgR, imgL)  # .astype(np.float32)/16
    displ = np.int16(displ)
    dispr = np.int16(dispr)

    filtered_disp = wls_filter.filter(displ, imgL, None, dispr)  # 对视差图进行滤波处理
    filtered_disp = cv2.normalize(
        src=filtered_disp, dst=None, beta=0, alpha=255, norm_type=cv2.NORM_MINMAX)
    filtered_disp = np.uint8(filtered_disp)

    # 保存深度图
    cv2.imwrite('depth_map.png', filtered_disp)

    return filtered_disp