#print height, " ", width #sleep(5) #gray = cv2.resize(image, dim, interpolation = cv2.INTER_AREA) #height, width = gray.shape[:2] # Our operations on the frame come here, doing some manipulation to reduce effects from surrounding objects, #light&shadow play, etc gray = cv2.medianBlur(gray, 5) #gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY) gray = cv2.bilateralFilter(gray, 5, 25, 25) #cv2.rectangle(gray, (width/2-150, height/2-150), (width/2+200, height/2+150), (255,0,0), 2) cv2.imshow('frame',gray) #showing image #gray = gray.copy()[height/2-150:height/2+150,width/2-150:width/2+200] image = Image.fromarray(gray) #turning array into image speech = ocr.ocr_image(image) #ocr-ing image by Tesseract #print type(speech) #speech = ocr.get_utf8_text() if ocr.get_mean_confidence() >= 63 and len(speech.strip()) >1: tts.say(str(speech)) print speech sleep(1) #robot speech if cv2.waitKey(1) & 0xFF == ord('q'): break # When everything done, release the capture #cap.release() camProxy.unsubscribe(videoClient) cv2.destroyAllWindows()
cv2.imshow('frame',gray) #showing image if cv2.waitKey(1) & 0xFF == ord('p'): cv2.imwrite('pic{:>05}.jpg'.format(i), gray) i += 1 #gray = gray.copy()[height/2-150:height/2+150,width/2-150:width/2+200] image = Image.fromarray(gray) #turning array into image #img = cv2.QueryFrame(image) speech = ocr.ocr_image(image) #ocr-ing image by Tesseract #print type(speech) #speech = ocr.get_utf8_text() wordList = re.sub("[^\w]", " ", str(speech)).split() if ocr.get_mean_confidence() >= 63: for word in wordList: print word # print dict.dictionary if dict.find_word(word.strip()): tts.say(word) print " the word is " + word # tts.say(str(speech)) # print speech sleep(1) #robot speech if cv2.waitKey(1) & 0xFF == ord('q'): break # When everything done, release the capture