Last active
September 10, 2017 17:24
-
-
Save shinob/b40e2a26fbd7493e784a9e2f24129017 to your computer and use it in GitHub Desktop.
OpenCVを使ったパターンマッチングで画像中の物体抽出 with Python ref: http://qiita.com/mix_dvd/items/b2b50ced80be1dcdbf2e
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
import cv2 | |
import math |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
from IPython.display import display, Image | |
def display_cv_image(image, format='.png'): | |
decoded_bytes = cv2.imencode(format, image)[1].tobytes() | |
display(Image(data=decoded_bytes)) |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
# 画像読込 | |
img1 = cv2.imread("IMG_4777.JPG") | |
img2 = cv2.imread("IMG_4754s.JPG") | |
# A-KAZE検出器の生成 | |
detector = cv2.AKAZE_create() | |
# 特徴量の検出と特徴量ベクトルの計算 | |
kp1, des1 = detector.detectAndCompute(img1, None) | |
kp2, des2 = detector.detectAndCompute(img2, None) | |
# Brute-Force Matcherの生成 | |
bf = cv2.BFMatcher() | |
# 特徴量ベクトル同士をBrute-Force&KNNでマッチング | |
matches = bf.knnMatch(des1, des2, k=2) | |
# データを間引く | |
ratio = 0.2 | |
good = [] | |
for m, n in matches: | |
if m.distance < ratio * n.distance: | |
good.append([m]) | |
# 特徴量をマッチング状況に応じてソート | |
good = sorted(matches, key = lambda x : x[1].distance) | |
# 対応する特徴点同士を描画 | |
img3 = cv2.drawMatchesKnn(img1, kp1, img2, kp2, good[:2], None, flags=2) | |
display_cv_image(img3, '.png') |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
# 特徴量データを取得 | |
q_kp = [] | |
t_kp = [] | |
for p in good[:2]: | |
for px in p: | |
q_kp.append(kp1[px.queryIdx]) | |
t_kp.append(kp2[px.trainIdx]) | |
# 加工対象の画像から特徴点間の角度と距離を計算 | |
q_x1, q_y1 = q_kp[0] | |
q_x2, q_y2 = q_kp[-1] | |
q_deg = math.atan2(q_y2 - q_y1, q_x2 - q_x1) * 180 / math.pi | |
q_len = math.sqrt((q_x2 - q_x1) ** 2 + (q_y2 - q_y1) ** 2) | |
# テンプレート画像から特徴点間の角度と距離を計算 | |
t_x1, t_y1 = t_kp[0] | |
t_x2, t_y2 = t_kp[-1] | |
t_deg = math.atan2(t_y2 - t_y1, t_x2 - t_x1) * 180 / math.pi | |
t_len = math.sqrt((t_x2 - t_x1) ** 2 + (t_y2 - t_y1) ** 2) | |
# 切出し位置の計算 | |
x1 = q_x1 - t_x1 * (q_len / t_len) | |
x2 = x1 + img2.shape[1] * (q_len / t_len) | |
y1 = q_y1 - t_y1 * (q_len / t_len) | |
y2 = y1 + img2.shape[0] * (q_len / t_len) | |
# 画像サイズ | |
x, y, c = img1.shape | |
size = (x, y) | |
# 回転の中心位置 | |
center = (q_x1, q_y1) | |
# 回転角度 | |
angle = q_deg - t_deg | |
# サイズ比率 | |
scale = 1.0 | |
# 回転変換行列の算出 | |
rotation_matrix = cv2.getRotationMatrix2D(center, angle, scale) | |
# アフィン変換 | |
img_rot = cv2.warpAffine(img1, rotation_matrix, size, flags=cv2.INTER_CUBIC) | |
# 画像の切出し | |
img_rot = img_rot[y1:y2, x1:x2] | |
# 縮尺調整 | |
x, y, c = img2.shape | |
img_rot = cv2.resize(img_rot, (y, x)) | |
# 結果表示 | |
display_cv_image(img_rot, '.png') |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment