人脸识别为什么用python开发

微信公开课小程序热门讨论「人脸识别为什么用python开发」最新回复-可以使用OpenCV，OpenCV的人脸检测功能在一般场合还是不错的。而ubuntu正好提供了pythonopencv这个包，用它可以方便地实现人脸检测的代码。写代码之前应该先安装pythonopencv：#!/usr/bin/python#*coding:UTF8*#face_detect.py#FaceDetectionusingOpenCV.Basedonsamplecodefrom:#http://python.pastebin.com/m76db1d6b#Usage:pythonface_detect.pyimage_fileimportsys,osfromopencv.cvimport*fromopencv.highguiimport*fromPILimportImage,ImageDrawfrommathimportsqrtdefdetectObjects(image):Convertsanimagetograyscaleandprintsthelocationsofanyfacesfoundgrayscale=cvCreateImage(cvSize(image.width,image.height),8,1)cvCvtColor(image,grayscale,CV_BGR2GRAY)storage=cvCreateMemStorage(0)cvClearMemStorage(storage)cvEqualizeHist(grayscale,grayscale)cascade=cvLoadHaarClassifierCascade(/usr/share/opencv/haarcascades/haarcascade_frontalface_default.xml,cvSize(1,1))faces=cvHaarDetectObjects(grayscale,cascade,storage,1.1,2,CV_HAAR_DO_CANNY_PRUNING,cvSize(20,20))result=[]forfinfaces:result.append((f.x,f.y,f.x+f.width,f.y+f.height))returnresultdefgrayscale(r,g,b):returnint(r*.3+g*.59+b*.11)defprocess(infile,outfile):image=cvLoadImage(infile);ifimage:faces=detectObjects(image)im=Image.open(infile)iffaces:draw=ImageDraw.Draw(im)forfinfaces:draw.rectangle(f,outline=(255,0,255))im.save(outfile,JPEG,quality=100)else:printError:cannotdetectfaceson%s%infileif__name__==__main__:process(input.jpg,output.jpg)
就这一个功能没多少钱，但app是个功能的集合，一般都会含有多种功能一起，像你这样的app一般是4到5万差不多。app广告能卖多少钱，看你的app受欢迎程度还有你和广告商具体怎么谈了 更多有关「人脸识别为什么用python开发」的疑问请扫码关注微信公开课+小程序！了解更多