图46.3.2.1 人脸识别实验流程图
46.3.3 main.py代码
main.py中的脚本代码如下所示：
import lcd
import sensor
import gc
from maix import KPU
lcd.init()
sensor.reset()
sensor.set_framesize(sensor.QVGA)
sensor.set_pixformat(sensor.RGB565)
sensor.set_hmirror(False)
anchor = (8.30891522166988, 2.75630994889035, 5.18609903718768, 1.7863757404970702, 6.91480529053198, 3.825771881004435, 10.218567655549439, 3.69476690620971, 6.4088204258368195, 2.38813526350986)
names = []
# 构造并初始化车牌检测KPU对象
lp_detecter = KPU()
lp_detecter.load_kmodel('/sd/KPU/lp_detect.kmodel')
lp_detecter.init_yolo2(anchor, anchor_num=len(anchor) // 2, img_w=320, img_h=240, net_w=320, net_h=240, layer_w=20, layer_h=15, threshold=0.7, nms_value=0.3, classes=len(names))
provinces = ['Wan', 'Hu', 'Jin', 'Yu^', 'Ji', 'Sx', 'Meng', 'Liao', 'Jl', 'Hei', 'Su', 'Zhe', 'Jing', 'Min', 'Gan', 'Lu', 'Yu', 'E^', 'Xiang', 'Yue', 'Gui^', 'Qiong', 'Cuan', 'Gui', 'Yun', 'Zang', 'Shan', 'Gan^', 'Qing', 'Ning', 'Xin']
ads = ['A', 'B', 'C', 'D', 'E', 'F', 'G', 'H', 'J', 'K', 'L', 'M', 'N', 'P', 'Q', 'R', 'S', 'T', 'U', 'V', 'W', 'X', 'Y', 'Z', '0', '1', '2', '3', '4', '5', '6', '7', '8', '9']
# 构造并初始化车牌识别KPU对象
lp_recognizer = KPU()
lp_recognizer.load_kmodel("/sd/KPU/lp_recog.kmodel")
lp_recognizer.lp_recog_load_weight_data("/sd/KPU/lp_weight.bin")
# 按指定比例扩展矩形框
def extend_box(x, y, w, h, scale):
x1 = int(x - scale * w)
x2 = int(x + w - 1 + scale * w)
y1 = int(y - scale * h)
y2 = int(y + h - 1 + scale * h)
x1 = x1 if x1 > 0 else 0
x2 = x2 if x2 < (320 - 1) else (320 - 1)
y1 = y1 if y1 > 0 else 0
y2 = y2 if y2 < (240 - 1) else (240 - 1)
return x1, y1, x2 - x1 + 1, y2 - y1 + 1
while True:
img = sensor.snapshot()
lp_detecter.run_with_output(img)
lps = lp_detecter.regionlayer_yolo2()
for lp in lps:
# 框出车牌位置
x, y, w, h = extend_box(lp[0], lp[1], lp[2], lp[3], 0.08)
img.draw_rectangle(x, y, w, h, color=(0, 255, 0))
# 对车牌进行车牌识别并绘制识别结果
lp = []
lp_img = img.cut(x, y, w, h)
resize_img = lp_img.resize(208, 64)
resize_img.replace(hmirror=True)
resize_img.pix_to_ai()
lp_recognizer.run_with_output(resize_img)
output = lp_recognizer.lp_recog()
for o in output:
lp.append(o.index(max(o)))
img.draw_string(x + 2, y - 20 - 2, '%s %s-%s%s%s%s%s' %(provinces[lp[0]], ads[lp[1]], ads[lp[2]], ads[lp[3]], ads[lp[4]], ads[lp[5]], ads[lp[6]]), color=(255, 0, 0), scale=2)
del lp
del lp_img
del resize_img
lcd.display(img)
gc.collect()
可以看到一开始是先初始化了LCD和摄像头，并分别构造并初始化了用于车牌检测和车牌号识别的KPU对象。
然后便是在一个循环中不断地获取摄像头输出的图像，首先将图像进行车牌检测，检测图像中存在的车牌，接着对车牌图像进行车牌号识别，最后将检测到的车牌位置框和识别出的车牌号码在图像上进行绘制，然后在LCD上显示图像。
46.4 运行验证
将DNK210开发板连接CanMV IDE，点击CanMV IDE上的“开始(运行脚本)”按钮后，将摄像头对准车牌，让其采集到车牌图像，可以看到LCD上显示了车牌识别的结果，图像中的被检测到的车牌均被框出，并且显示了车牌对应识别出的车牌号码，如下图所示： 
0
/7 
工商网监
湘ICP备2023018690号
225
淘帖
