nms【非极大值抑制】的原理+python代码实现

参考链接：

nms原理

http://www.cnblogs.com/liekkas0626/p/5219244.html

python代码实现

https://blog.csdn.net/hongxingabc/article/details/78996407

 

额外链接：

Non-Maximum Suppression for Object Detection in Python: http://pyimg.co/fz1ak
• (Faster) Non-Maximum Suppression in Python: http://pyimg.co/gwunq
I would also recommend looking at my non-maxima suppression implementation in the
imutils library which we will be using later in this chapter: http://pyimg.co/9dne5
 

-----------------------------------------------------------------------------

 

非极大值抑制算法

1. 算法原理

  非极大值抑制算法（Non-maximum suppression, NMS）的本质是搜索局部极大值，抑制非极大值元素。

2. 3邻域情况下NMS的实现

  3邻域情况下的NMS即判断一维数组I[W]的元素I[i](2<=i<=W-1)是否大于其左邻元素I[i-1]和右邻元素I[i+1]，算法流程如下图所示：

  

  a. 算法流程3-5行判断当前元素是否大于其左邻与右邻元素，如符合条件，该元素即为极大值点。对于极大值点I[i]，已知I[i]>I[i+1]，故无需对i+1位置元素做进一步处理，直接跳至i+2位置，对应算法流程第12行。

    

  b. 若元素I[i]不满足算法流程第3行判断条件，将其右邻I[i+1]作为极大值候选，对应算法流程第7行。采用单调递增的方式向右查找，直至找到满足I[i]>I[i+1]的元素，若i<=W-1，该点即为极大值点，对应算法流程第10-11行。

    

3. NMS在物体检测中的应用

  物体检测中应用NMS算法的主要目的是消除多余（交叉重复）的窗口，找到最佳物体检测位置。

  

  如上图所示，人脸检测中，虽然每个窗口均检测到人脸，但仅需给出一个最有可能表征人脸的窗口。

=============================================

基于Python的代码实现：

 

 

目标检测中常用到NMS，在faster R-CNN中，每一个bounding box都有一个打分，NMS实现逻辑是：

1，按打分最高到最低将BBox排序 ，例如：A B C D E F

2，A的分数最高，保留。从B-E与A分别求重叠率IoU，假设B、D与A的IoU大于阈值，那么B和D可以认为是重复标记去除

3，余下C E F，重复前面两步。

 

#coding:utf-8  
import numpy as np    
def py_cpu_nms(dets, thresh):  
    """Pure Python NMS baseline."""  
    x1 = dets[:, 0]  
    y1 = dets[:, 1]  
    x2 = dets[:, 2]  
    y2 = dets[:, 3]  
    scores = dets[:, 4]  #bbox打分  
    areas = (x2 - x1 + 1) * (y2 - y1 + 1)  
#打分从大到小排列，取index  
    order = scores.argsort()[::-1]  
#keep为最后保留的边框  
    keep = []  
    while order.size > 0:  
#order[0]是当前分数最大的窗口，肯定保留  
        i = order[0]  
        keep.append(i)  
#计算窗口i与其他所有窗口的交叠部分的面积
        xx1 = np.maximum(x1[i], x1[order[1:]])  
        yy1 = np.maximum(y1[i], y1[order[1:]])  
        xx2 = np.minimum(x2[i], x2[order[1:]])  
        yy2 = np.minimum(y2[i], y2[order[1:]])   
        w = np.maximum(0.0, xx2 - xx1 + 1)  
        h = np.maximum(0.0, yy2 - yy1 + 1)  
        inter = w * h  
#交/并得到iou值  
        ovr = inter / (areas[i] + areas[order[1:]] - inter)  
#inds为所有与窗口i的iou值<=threshold值的窗口的index，其他窗口也就是iou>threshold的值因为重复了所以被删除，仅仅保留iou<=threshold
        inds = np.where(ovr <= thresh)[0]  
#order里面只保留与窗口i交叠面积小于threshold的那些窗口，由于ovr长度比order长度少1(不包含i)，所以inds+1对应到保留的窗口
        order = order[inds + 1]    
    return keep

=============================

另一个版本的nms代码，比较好理解一些

# import the necessary packages
import numpy as np

def non_max_suppression(boxes, probs=None, overlapThresh=0.3):
	# if there are no boxes, return an empty list
	if len(boxes) == 0:
		return []

	# if the bounding boxes are integers, convert them to floats -- this
	# is important since we'll be doing a bunch of divisions
	if boxes.dtype.kind == "i":
		boxes = boxes.astype("float")

	# initialize the list of picked indexes
	pick = []

	# grab the coordinates of the bounding boxes
	x1 = boxes[:, 0]
	y1 = boxes[:, 1]
	x2 = boxes[:, 2]
	y2 = boxes[:, 3]

	# compute the area of the bounding boxes and grab the indexes to sort
	# (in the case that no probabilities are provided, simply sort on the
	# bottom-left y-coordinate)
	area = (x2 - x1 + 1) * (y2 - y1 + 1)
	idxs = y2

	# if probabilities are provided, sort on them instead
	if probs is not None:
		idxs = probs

	# sort the indexes
	idxs = np.argsort(idxs)

	# keep looping while some indexes still remain in the indexes list
	while len(idxs) > 0:
		# grab the last index in the indexes list and add the index value
		# to the list of picked indexes
		last = len(idxs) - 1
		i = idxs[last]
		pick.append(i)

		# find the largest (x, y) coordinates for the start of the bounding
		# box and the smallest (x, y) coordinates for the end of the bounding
		# box
		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]])

		# compute the width and height of the bounding box
		w = np.maximum(0, xx2 - xx1 + 1)
		h = np.maximum(0, yy2 - yy1 + 1)

		# compute the ratio of overlap
		overlap = (w * h) / area[idxs[:last]]

		# delete all indexes from the index list that have overlap greater
		# than the provided overlap threshold
		idxs = np.delete(idxs, np.concatenate(([last],
			np.where(overlap > overlapThresh)[0])))

	# return only the bounding boxes that were picked
	return boxes[pick].astype("int")