제로베이스 데이터취업스쿨 DAY84 OpenCV5~8

명함 검출과 인식

src = cv2.imread('/Users/nayoung/Documents/ds_study/data/namecard1.jpg')

if src is None:
    print('Image load failed!')
    sys.exit()

# 그레이스케일로 변환
src_gray = cv2.cvtColor(src, cv2.COLOR_BGR2GRAY)

# 자동 이진화
th, src_bin = cv2.threshold(src_gray, 0, 255, cv2.THRESH_BINARY | cv2.THRESH_OTSU)

# 외곽선 검출
contours, _ = cv2.findContours(src_bin, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_NONE)

'''
# 이진 영상을 컬러 영상 형식으로 변환
dst = cv2.cvtColor(src_bin, cv2.COLOR_GRAY2BGR)

# 검출된 외곽선 모두 그리기
for i in range(len(contours)):
    c = (random.randint(0,255), random.randint(0,255), random.randint(0,255))
    cv2.drawContours(dst, contours, i, c, 2)  # 2: 픽셀 두께

cv2.imshow('src', src)
cv2.imshow('dst', dst)
cv2.waitKey()
cv2.destroyAllWindows()
'''

for pts in contours:
    # 너무 작은 객체는 무시
    if cv2.contourArea(pts) < 1000:
        continue

    # 외곽선 근사화
    approx = cv2.approxPolyDP(pts, cv2.arcLength(pts, True)*0.02, True)

    # 사각형으로 근사화되면 외곽선 표시
    if len(approx) == 4:
        cv2.polylines(src, [approx], True, (0,255,0), 2, cv2.LINE_AA)

cv2.imshow('src', src)
cv2.waitKey()
cv2.destroyAllWindows()

명함 똑바로 피기1

src = cv2.imread('/Users/nayoung/Documents/ds_study/data/namecard1.jpg')

w, h = 720, 400
srcQuad = np.array([[324,308], [760,369], [718,611], [231,517]], np.float32)
dstQuad = np.array([[0,0], [w-1,0], [w-1,h-1], [0,h-1]], np.float32)
dst = np.zeros((h,w), np.uint8)

pers = cv2.getPerspectiveTransform(srcQuad, dstQuad)
dst = cv2.warpPerspective(src, pers, (w,h))

cv2.imshow('src', src)
cv2.imshow('dst', dst)
cv2.waitKey()
cv2.destroyAllWindows()

명함 똑바로 피기2

def reorderPts(pts):
    idx = np.lexsort((pts[:,1], pts[:,0]))
    pts = pts[idx]

    if pts[0,1] > pts[1,1]:
        pts[[0,1]] = pts[[1,0]]
   
    if pts[2,1] < pts[3,1]:
        pts[[2,3]] = pts[[3,2]]

    return pts

src = cv2.imread('/Users/nayoung/Documents/ds_study/data/namecard1.jpg')

# 출력 영상 설정
w, h = 720, 400
srcQuad = np.array([[0,0], [0,h], [w,h], [w,0]], np.float32)
dstQuad = np.array([[0,0], [0,h], [w,h], [w,0]], np.float32)
dst = np.zeros((h,w), np.uint8)

# 입력 영상을 그레이스케일로 변환
src_gray = cv2.cvtColor(src, cv2.COLOR_BGR2GRAY)

# 자동 이진화
_, src_bin = cv2.threshold(src_gray, 0, 255, cv2.THRESH_BINARY | cv2.THRESH_OTSU)

# 외곽선 검출 및 명함 검출
contours, _ = cv2.findContours(src_bin, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_NONE)

for pts in contours:
    # 너무 작은 객체는 무시
    if cv2.contourArea(pts) < 1000:
        continue

    # 외곽선 근사화
    approx = cv2.approxPolyDP(pts, cv2.arcLength(pts, True)*0.02, True)

    # 사각형으로 근사화되면 외곽선 표시
    if len(approx) == 4:
        # cv2.polylines(src, [approx], True, (0,255,0), 2, cv2.LINE_AA)
        corners = approx.reshape(4,2).astype(np.float32)
        srcQuad = reorderPts(corners)

pers = cv2.getPerspectiveTransform(srcQuad, dstQuad)
dst = cv2.warpPerspective(src, pers, (w,h))

cv2.imshow('src', src)
cv2.imshow('dst', dst)
cv2.waitKey()
cv2.destroyAllWindows()

얼굴 검출과 응용

얼굴 검출

import sys
import numpy as np
import cv2


model = '/Users/nayoung/Documents/ds_study/data/res10_300x300_ssd_iter_140000_fp16.caffemodel'
config = '/Users/nayoung/Documents/ds_study/data/deploy.prototxt.txt'
#model = 'opencv_face_detector/opencv_face_detector_uint8.pb'
#config = 'opencv_face_detector/opencv_face_detector.pbtxt'

cap = cv2.VideoCapture(0)

if not cap.isOpened():
    print('Camera open failed!')
    sys.exit()

net = cv2.dnn.readNet(model, config)

if net.empty():
    print('Net open failed!')
    sys.exit()

while True:
    ret, frame = cap.read()

    if not ret:
        break

    blob = cv2.dnn.blobFromImage(frame, 1, (300, 300), (104, 177, 123))
    net.setInput(blob)

    out = net.forward()
    detect = out[0, 0, :, :]

    (h, w) = frame.shape[:2]

    for i in range(detect.shape[0]):
        confidence = detect[i, 2]
        if confidence < 0.5:
            break

        x1 = int(detect[i, 3] * w)
        y1 = int(detect[i, 4] * h)
        x2 = int(detect[i, 5] * w)
        y2 = int(detect[i, 6] * h)

        cv2.rectangle(frame, (x1, y1), (x2, y2), (0, 255, 0))

        label = f'Face: {confidence:4.2f}'
        cv2.putText(frame, label, (x1, y1 - 1), cv2.FONT_HERSHEY_SIMPLEX, 0.8, (0, 255, 0), 1, cv2.LINE_AA)

    cv2.imshow('frame', frame)

    if cv2.waitKey(1) == 27:
        break

cap.release()
cv2.destroyAllWindows()

모자이크

import sys
import numpy as np
import cv2


model = '/Users/nayoung/Documents/ds_study/data/res10_300x300_ssd_iter_140000_fp16.caffemodel'
config = '/Users/nayoung/Documents/ds_study/data/deploy.prototxt.txt'
#model = 'opencv_face_detector/opencv_face_detector_uint8.pb'
#config = 'opencv_face_detector/opencv_face_detector.pbtxt'

cap = cv2.VideoCapture(0)

if not cap.isOpened():
    print('Camera open failed!')
    sys.exit()

net = cv2.dnn.readNet(model, config)

if net.empty():
    print('Net open failed!')
    sys.exit()

while True:
    ret, frame = cap.read()

    if not ret:
        break

    blob = cv2.dnn.blobFromImage(frame, 1, (300, 300), (104, 177, 123))
    net.setInput(blob)

    out = net.forward()
    detect = out[0, 0, :, :]

    (h, w) = frame.shape[:2]

    for i in range(detect.shape[0]):
        confidence = detect[i, 2]
        if confidence < 0.5:
            break

        x1 = int(detect[i, 3] * w)
        y1 = int(detect[i, 4] * h)
        x2 = int(detect[i, 5] * w)
        y2 = int(detect[i, 6] * h)

        #cv2.rectangle(frame, (x1, y1), (x2, y2), (0, 255, 0), 2)

        face_img = frame[y1:y2, x1:x2]
        fh, fw = face_img.shape[:2]

        face_img2 = cv2.resize(face_img, (0, 0), fx=1./16, fy=1./16)
        cv2.resize(face_img2, (fw, fh), dst=face_img, interpolation=cv2.INTER_NEAREST)
        #frame[y1:y2, x1:x2] = cv2.resize(face_img2, (fw, fh), interpolation=cv2.INTER_NEAREST)

    cv2.imshow('frame', frame)

    if cv2.waitKey(1) == 27:
        break

cap.release()
cv2.destroyAllWindows()