Darknet-修改源码使用python文件调用模型检测视频流

关于yolo的介绍、发展及使用，网上有很多资料，这里就先不进行赘述了，
yolo（darknet）主页：https://pjreddie.com/darknet/yolo/

一、下载源码
（此处我使用的源码不是最新的，使用的源码可以在我上传的资源中找到https://download.csdn.net/download/z_muyangren/10965702）
git clone https://github.com/pjreddie/darknet

二、修改源码

===================
darknet.h
===================
#ifndef __cplusplus
#ifdef OPENCV
image get_image_from_stream(CvCapture *cap);
//---------------------------------------------------------------------------
image load_image_cap(unsigned int *matrix, int rows, int cols, int channels);
//---------------------------------------------------------------------------
#endif
#endif
=========================
image.c
=========================
#ifdef OPENCV
image load_image_cap(unsigned int *matrix, int rows, int cols, int channels)
{
	IplImage *src;
	src = cvCreateImageHeader(cvSize(cols,rows),IPL_DEPTH_8U,channels);
	int size = rows*cols*3;
	unsigned char imageData[size];
	int a = 0;
	int b = 0;
	for (a=0; a<rows*cols*2*3; a++)
	{
		if (0==a%2)
		{
			imageData[b] = (unsigned char)matrix[a];
			b++;
		}
	}
	//stepnum: the bytes number of every row
	int stepnum = cols * 3;
	cvSetData(src,imageData,stepnum);

    int h = src->height;
    int w = src->width;
    int c = src->nChannels;
    image out;
    out.data = 0;
    out.h = h;
    out.w = w;
    out.c = c;
    out.data = calloc(h*w*c, sizeof(float));
    unsigned char *data = (unsigned char *)src->imageData;
	int step = src->widthStep;

	int i, j, k;
	for(i = 0; i < h; ++i){
		for(k= 0; k < c; ++k){
			for(j = 0; j < w; ++j){
				out.data[k*w*h + i*w + j] = data[i*step + j*c + k]/255.;
			}
		}
	}
    return out;
}
#endif

三、编译
修改Makefile文件 GPU=1 CUDNN=1 OPENCV=1

四、数据处理、训练模型（此处直接使用yolov3.weights）

五、修改python文件

'''
修改：
#1
load_image_cap = lib.load_image_cap
load_image_cap.argtypes = [c_char_p, c_int, c_int, c_int]
load_image_cap.restype = IMAGE
#2
cols= image.shape[1]
rows =image.shape[0]
channels= 3
#need have the same type with int* img
image= np.asarray(image, dtype=np.int)
#change a multidimensional array to char*
src = image.ctypes.data_as(ctypes.c_char_p)
im = load_image_cap(src, rows, cols, channels)
#3
set_gpu = lib.cuda_set_device
set_gpu.argtypes = [c_int]
'''
from ctypes import *
import numpy as np
import math
import random
import cv2
import time
import ctypes

def sample(probs):
    s = sum(probs)
    probs = [a/s for a in probs]
    r = random.uniform(0, 1)
    for i in range(len(probs)):
        r = r - probs[i]
        if r <= 0:
            return i
    return len(probs)-1

def c_array(ctype, values):
    arr = (ctype*len(values))()
    arr[:] = values
    return arr

class BOX(Structure):
    _fields_ = [("x", c_float),
                ("y", c_float),
                ("w", c_float),
                ("h", c_float)]

class DETECTION(Structure):
    _fields_ = [("bbox", BOX),
                ("classes", c_int),
                ("prob", POINTER(c_float)),
                ("mask", POINTER(c_float)),
                ("objectness", c_float),
                ("sort_class", c_int)]


class IMAGE(Structure):
    _fields_ = [("w", c_int),
                ("h", c_int),
                ("c", c_int),
                ("data", POINTER(c_float))]

class METADATA(Structure):
    _fields_ = [("classes", c_int),
                ("names", POINTER(c_char_p))]

    

lib = CDLL("/home/muyangren/darknet/libdarknet.so", RTLD_GLOBAL)
#lib = CDLL("libdarknet.so", RTLD_GLOBAL)
lib.network_width.argtypes = [c_void_p]
lib.network_width.restype = c_int
lib.network_height.argtypes = [c_void_p]
lib.network_height.restype = c_int

predict = lib.network_predict
predict.argtypes = [c_void_p, POINTER(c_float)]
predict.restype = POINTER(c_float)

set_gpu = lib.cuda_set_device
set_gpu.argtypes = [c_int]

make_image = lib.make_image
make_image.argtypes = [c_int, c_int, c_int]
make_image.restype = IMAGE

get_network_boxes = lib.get_network_boxes
get_network_boxes.argtypes = [c_void_p, c_int, c_int, c_float, c_float, POINTER(c_int), c_int, POINTER(c_int)]
get_network_boxes.restype = POINTER(DETECTION)

make_network_boxes = lib.make_network_boxes
make_network_boxes.argtypes = [c_void_p]
make_network_boxes.restype = POINTER(DETECTION)

free_detections = lib.free_detections
free_detections.argtypes = [POINTER(DETECTION), c_int]

free_ptrs = lib.free_ptrs
free_ptrs.argtypes = [POINTER(c_void_p), c_int]

network_predict = lib.network_predict
network_predict.argtypes = [c_void_p, POINTER(c_float)]

reset_rnn = lib.reset_rnn
reset_rnn.argtypes = [c_void_p]

load_net = lib.load_network
load_net.argtypes = [c_char_p, c_char_p, c_int]
load_net.restype = c_void_p

do_nms_obj = lib.do_nms_obj
do_nms_obj.argtypes = [POINTER(DETECTION), c_int, c_int, c_float]

do_nms_sort = lib.do_nms_sort
do_nms_sort.argtypes = [POINTER(DETECTION), c_int, c_int, c_float]

free_image = lib.free_image
free_image.argtypes = [IMAGE]

letterbox_image = lib.letterbox_image
letterbox_image.argtypes = [IMAGE, c_int, c_int]
letterbox_image.restype = IMAGE

load_meta = lib.get_metadata
lib.get_metadata.argtypes = [c_char_p]
lib.get_metadata.restype = METADATA

load_image = lib.load_image_color
load_image.argtypes = [c_char_p, c_int, c_int]
load_image.restype = IMAGE

load_image_cap = lib.load_image_cap
load_image_cap.argtypes = [c_char_p, c_int, c_int, c_int]
load_image_cap.restype = IMAGE

rgbgr_image = lib.rgbgr_image
rgbgr_image.argtypes = [IMAGE]

predict_image = lib.network_predict_image
predict_image.argtypes = [c_void_p, IMAGE]
predict_image.restype = POINTER(c_float)

def classify(net, meta, im):
    out = predict_image(net, im)
    res = []
    for i in range(meta.classes):
        res.append((meta.names[i], out[i]))
    res = sorted(res, key=lambda x: -x[1])
    return res

def detect(net, meta, image, thresh=.5, hier_thresh=.5, nms=.45):
    #im = load_image(image, 0, 0)
    cols= image.shape[1]
    rows =image.shape[0]
    channels= 3
    #need have the same type with int* img
    image= np.asarray(image, dtype=np.int)
    #change a multidimensional array to char*
    src = image.ctypes.data_as(ctypes.c_char_p)
    im = load_image_cap(src, rows, cols, channels)
    
    num = c_int(0)
    pnum = pointer(num)
    predict_image(net, im)
    dets = get_network_boxes(net, im.w, im.h, thresh, hier_thresh, None, 0, pnum)
    num = pnum[0]
    if (nms): do_nms_obj(dets, num, meta.classes, nms);

    res = []
    for j in range(num):
        for i in range(meta.classes):
            if dets[j].prob[i] > 0:
                b = dets[j].bbox
                res.append((meta.names[i], dets[j].prob[i], (b.x, b.y, b.w, b.h)))
    res = sorted(res, key=lambda x: -x[1])
    free_image(im)
    free_detections(dets, num)
    return res

if __name__ == "__main__":
    net = load_net("cfg/yolov3.cfg", "yolov3.weights", 0)
    meta = load_meta("/home/muyangren/darknet/cfg/coco.data")

    #read video frame
    #videoCapture = cv2.VideoCapture("/home/muyangren/darknet/59cbb31588870.mp4")
    videoCapture = cv2.VideoCapture(201)
    success = True
    while(videoCapture.isOpened()):
        success, frame = videoCapture.read()
        r = detect(net, meta, frame)

        print len(r)
        for i in range(len(r)):
	    box_left = (int)(r[i][2][0] - r[i][2][2] / 2.0)
            box_top = (int)(r[i][2][1] - r[i][2][3] / 2.0)
	    box_right = (int)(r[i][2][0] + r[i][2][2] / 2.0)
	    box_bottom = (int)(r[i][2][1] + r[i][2][3] / 2.0)  
	
            display_image = frame
            #draw the rectangle on the image.  This is hopefully around the object
            box_color = (0, 255, 0)  # green box
            box_thickness = 2
            cv2.rectangle(display_image, (box_left, box_top),(box_right, box_bottom), box_color, box_thickness)

            # draw the classification label string just above and to the left of the rectangle
            label_background_color = (70, 120, 70) # greyish green background for text
            label_text_color = (255, 255, 255)   # white text
            cv2.rectangle(display_image,(box_left, box_top-20),(box_right,box_top), label_background_color, -1)
            cv2.putText(display_image, r[i][0] + ' : %.2f' % r[i][1], (box_left+5,box_top-7), cv2.FONT_HERSHEY_SIMPLEX, 0.5, label_text_color, 1)

	print r
	cv2.imshow("result_window", display_image)
        time.sleep(0.1)

        if cv2.waitKey(1) == ord('q'):
            break
    capture.release()
    cv2.destroyAllWindows()
    

六、运行python文件，结果如下：