class trainLimitFrame(object): def __init__(self,framelength=100,label=0,xmlfile='./train/mytrainSet.xml'): self._framelen=framelength self._xmlfile=xmlfile self._label=label self._core=core(faceRecognizeXmlfile=self._xmlfile) self._iosteam=IOSteam() self._window=PreviewWindowManager('TrainWindows') self._window_faceZone=PreviewWindowManager('TrainWindows.FaceZone') def run(self): self._window.createWindow() self._window_faceZone.createWindow() self._iosteam.run() for i in range(self._framelen): logging.info('current frame: {}'.format(i)) frame=self._iosteam.frame self._core.train(frame,self._label) # self._core._face_detection(frame) whattoShow=self._core.displayFrame self._window.show(whattoShow) if self._core.faceSetofImageinSameFrameResized!=[]: self._window_faceZone.show(self._core.faceSetofImageinSameFrameResized[0]) cv2.waitKey(1) self._iosteam.stop() self._core.saveTrainSet() self._window.destroyWindow() self._window_faceZone.destroyWindow()
def __init__(self,framelength=100,label=0,xmlfile='./train/mytrainSet.xml'): self._framelen=framelength self._xmlfile=xmlfile self._label=label self._core=core(faceRecognizeXmlfile=self._xmlfile) self._iosteam=IOSteam() self._window=PreviewWindowManager('TrainWindows') self._window_faceZone=PreviewWindowManager('TrainWindows.FaceZone')
def __init__(self,xmlfile='./train/mytrainSet.xml',imgDB='faceDB'): self._xmlfile=xmlfile self._trainedImgPostion=imgDB self._imgMat=[] self._label=[] self._core=core(faceRecognizeXmlfile=self._xmlfile) self._iosteam=IOSteam() self._window=PreviewWindowManager('TrainWindows',self.onKeypress) self._window_faceZone=PreviewWindowManager('TrainWindows.FaceZone') self._trained_file_used=False self._keep_run=True self._apply_recognize=False self.train_setted_img()
class application(object): def __init__(self,xmlfile='./train/mytrainSet.xml',imgDB='faceDB'): self._xmlfile=xmlfile self._trainedImgPostion=imgDB self._imgMat=[] self._label=[] self._core=core(faceRecognizeXmlfile=self._xmlfile) self._iosteam=IOSteam() self._window=PreviewWindowManager('TrainWindows',self.onKeypress) self._window_faceZone=PreviewWindowManager('TrainWindows.FaceZone') self._trained_file_used=False self._keep_run=True self._apply_recognize=False self.train_setted_img() def _link_img_lable(self,img,label): self._imgMat.append(img) self._label.append(label) def _load_image_trained(self): oswalk=list(os.walk(self._trainedImgPostion)) for current_dir,sub_dir,sub_file in oswalk: if sub_file!=[]: for eachfile in sub_file: fileposition=current_dir+'\\'+eachfile img=cv2.imread(fileposition) label=current_dir.split('\\')[-1] self._link_img_lable(img,int(label)) logging.debug('file walker:filePosition:{} label:{}'.format( fileposition, label )) def onKeypress(self, keycode): """Handle a keypress. space -> Take a screenshot. escape -> Quit. r -> Reverse mode of faceRecognize,(default:off) m -> Move png file to faceSetDataBase,Please input an integer[1,2,3,4,5...] to tell the programe how to settle current *.png.after operation,png files will be moved to (./faceDB/[label]/i.png) """ if keycode == 32: # space self._iosteam.dumpFrame() elif keycode == 27: # escape #self._windowManager.destroyWindow() self._keep_run=False elif keycode == ord('r'): #activate face recognize self._apply_recognize=not self._apply_recognize logging.info('[+]app recognize mode change to {}'.format( self._apply_recognize)) if self._apply_recognize: #self.train_setted_img() pass elif keycode == ord('m'): #move file .png to label label=raw_input('[*]Please input the interger ID of current dirs\' photo:\n') try: os.mkdir('./faceDB/'+label) except WindowsError as e: logging.info('Directory error{},means that existed'.format(e)) filename=[i for i in os.listdir('.') if i.endswith('png')] logging.info('file:{}'.format(filename)) if 0==os.system('mv *.png ./faceDB/'+label): logging.info('[+]*.png moved Successfully') else: logging.info('[-]*.png moved Fail') def train_setted_img(self): """ run during the init of app,pre train gived image with label """ self._load_image_trained() for i in range(len(self._label)): self._core.train( self._imgMat[i],self._label[i] ) logging.debug('pre train step used img{} ,label{} '.format( self._imgMat[i].shape,self._label[i])) self._trained_file_used=True def _runwithIOSteam(self): self._window.createWindow() self._window_faceZone.createWindow() self._iosteam.run() while self._keep_run: frame=self._iosteam.frame if(frame==None): continue #face recognize if self._apply_recognize: self._core.recognize(frame) result=self._core.result else: #just detect faceZone self._core._face_detection(frame) #preview result here whattoShow=self._core.displayFrame if self._apply_recognize: self._draw_name_in_frame(whattoShow) self._window.show(whattoShow) if self._core.faceSetofImageinSameFrameResized!=[]: self._window_faceZone.show(self._core.faceSetofImageinSameFrameResized[0]) self._window.processEvents() self._iosteam.stop() self._window.destroyWindow() self._window_faceZone.destroyWindow() def _draw_name_in_frame(self,frame): for eachPeople in self._core.result: true_name=faceIDmatchName[eachPeople['who'][0]] dist=str(eachPeople['who'][1]) pos=tuple(eachPeople['zone'][:2]) draw_name(frame,true_name+dist,pos) def run(self): if(self._trained_file_used): self._runwithIOSteam() else: logging.debug('faceDB do not used in Train')
# -*- coding: utf-8 -*- """ Created on Sun Jun 19 17:33:35 2016 test FaceDetectation @author: pip """ import cv2 import logging from cameraIO import IOSteam import sys from trackers import FaceTracker front=IOSteam() tracker=FaceTracker() front.run() tracker.update( front.frame) facelist=tracker.faces front.stop() print(facelist)