测试与发现
YOLOv5官方给出的YOLOv5在OpenCV上推理的程序相对来说是比较通俗易懂的,条理清晰,有基本的封装,直接可用!但是我也发现,模型的推理时间跟前后处理的时间相差无几,特别是当视频流有多个检测到的对象时候,整个帧率会有明显下降!官方推荐的参考示例代码链接为:
https://github.com/doleron/yolov5-opencv-cpp-python/blob/main/python/yolo-tiny.py最后发现推理时间没有明显变化,主要是前后处理,有两个函数耗时比较高!从输入图像转换到模型输入数据的函数:
cv2.dnn.blobFromImage(input_image , 1/255.0, (640, 640), swapRB=True)推理之后的重叠目标框非最大抑制函数:
indexes = cv2.dnn.NMSBoxes(boxes, confidences, 0.25, 0.45)特别是非最大抑制函数,随着图像中目标数目增多,导致帧率成明显下降趋势!
修改输入转换
cv2.dnn.blobFromImage(input_image , 1/255.0, (640, 640), swapRB=True)
可以通过下面的代码等价替换:
rgb = cv.cvtColor(image, cv.COLOR_BGR2RGB) input_image = cv.resize(src=rgb, dsize=(INPUT_WIDTH, INPUT_HEIGHT)) blob_img = np.float32(input_image) / 255.0 input_x = blob_img.transpose((2, 0, 1)) input_blob = np.expand_dims(input_x, 0)
修改之后测试发现该替代降低了执行时间,说明替代有效!
修改非最大抑制
indexes = cv2.dnn.NMSBoxes(boxes, confidences, 0.25, 0.45)
输入的box格式x, y,w,h,我参考了网上的代码,修改实现一个基于并交比最简单的NMS抑制算法,基于矩阵计算,保证不会因为对象变得多了,增加计算耗时,然后把它们封装成一个单独的方法,导入该方法直接替换之前的代码行为:
class_ids, boxes = non_max_suppression_fast(np.asarray(class_ids), np.asarray(boxes), 0.75)
该函数完整的实现代码如下:
import numpy as np def non_max_suppression_fast(class_ids, boxes, nms_threshold): # if there are no boxes, return if len(boxes) == 0: return [], [] 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 sort the bounding # boxes by the bottom-right y-coordinate of the bounding box area = (x2 - x1 + 1) * (y2 - y1 + 1) idxs = np.argsort(y2) # 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 idxs = np.delete(idxs, np.concatenate(([last], np.where(overlap > nms_threshold)[0]))) # return only the bounding boxes that were picked using the # integer data type return class_ids[pick], boxes[pick].astype("int") if __name__ == "__main__": boxes = [] boxes.append((163, 0, 27+163, 41)) boxes.append((164, 0, 28+164, 43)) boxes.append((165, 0, 29+165, 42)) res = non_max_suppression_fast(None, np.asarray(boxes), 0.25) print(res)
对比测试
两处都修改完成之后,其它输入条件与代码不变,硬件相同条件下对比测试效果如下:修改之前 Python版本OpenCV与OpenVINO上推理速度:
修改之后 Python版本OpenCV与OpenVINO上推理速度:
可以看到FPS较之前有明显的提升!
全部0条评论
快来发表一下你的评论吧 !