|
在此基础上加上NMS算法，
NMS是经常伴随图像区域检测的算法，作用是去除重复的区域
"""
-*- coding: utf-8 -*-
author： Hao Hu
@date 2021/12/2 10:52 PM
"""
import cv2
img_path = 'sample1.jpg'
import numpy as np
def non_max_suppression_fast(boxes, overlapThresh):
"""将矩形框中的矩形框去掉"""
# 空数组检测
if len(boxes) == 0:
return []
# 将类型转为float
if boxes.dtype.kind == "i":
boxes = boxes.astype("float")
pick = []
# 四个坐标数组
x1 = boxes[:, 0]
y1 = boxes[:, 1]
x2 = boxes[:, 2]
y2 = boxes[:, 3]
area = (x2 - x1 + 1) * (y2 - y1 + 1) # 计算面积数组
idxs = np.argsort(y2) # 返回的是右下角坐标从小到大的索引值
# 开始遍历删除重复的框
while len(idxs) > 0:
# 将最右下方的框放入pick数组
last = len(idxs) - 1
i = idxs[last]
pick.append(i)
# 找到剩下的其余框中最大的坐标x1y1，和最小的坐标x2y2,
xx1 = np.maximum(x1[i], x1[idxs[:last]])
yy1 = np.maximum(y1[i], y1[idxs[:last]])
xx2 = np.minimum(x2[i], x2[idxs[:last]])
yy2 = np.minimum(y2[i], y2[idxs[:last]])
# 计算重叠面积占对应框的比例
w = np.maximum(0, xx2 - xx1 + 1)
h = np.maximum(0, yy2 - yy1 + 1)
overlap = (w * h) / area[idxs[:last]]
# 如果占比大于阈值，则删除
idxs = np.delete(idxs, np.concatenate(([last], np.where(overlap > overlapThresh)[0])))
return boxes[pick].astype("int")
def get_word_area(img_path):
"""得到检测图像中的文本区域，画出轮廓"""
mser = cv2.MSER_create()
img = cv2.imread(img_path)
gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
vis = img.copy()
regions, _ = mser.detectRegions(gray)
hulls = [cv2.convexHull(p.reshape(-1, 1, 2)) for p in regions]
# 绘制目前的矩形文本框
# mser = cv2.MSER_create()
# cv2.polylines(vis, hulls, 1, (0, 255, 0))
keep = []
for c in hulls:
x, y, w, h = cv2.boundingRect(c)
keep.append([x, y, x + w, y + h])
'''NMS是经常伴随图像区域检测的算法，作用是去除重复的区域，
在人脸识别、物体检测等领域都经常使用，全称是非极大值抑制（non maximum suppression），
就是抑制不是极大值的元素，所以用在这里就是抑制不是最大框的框，也就是去除大框中包含的小框'''
# 使用NMS算法
# keep2 = np.array(keep)
# pick = non_max_suppression_fast(keep2, 0.5)
# for (startX, startY, endX, endY) in pick:
# cv2.rectangle(vis, (startX, startY), (endX, endY), (255, 185, 120), 2)
# 直接使用holyholes算法
cv2.polylines(vis, hulls, 1, (0, 255, 0))
cv2.imshow("before use NMS", vis)
if cv2.waitKey(0) == 9:
cv2.destroyAllWindows()
if __name__ == '__main__':
img_path = 'sample1.jpg'
get_word_area(img_path)
|