CHEN, YEN LIN,陈彦霖,陳彥霖,YU, CHAO WEI,余兆伟,余兆偉,CHIANG, CHUAN YEN,江川彦,江川彥
申请号：
TW101103602
公开号：
TW201332539A
申请日:
2012.02.03
申请国别(地区):
TW
年份:
2013
代理人:
摘要：
This invention discloses a blind guiding system having the function of the real-time voice response and a method thereof. The system utilizes two processing modules to perform image processing and voice responding respectively so as to inform a user of front environmental conditions in time. The system comprises an image sensing module, a memory, an image processing module, a system processing module and an audio module. The image sensing module senses an image. The memory stores a plurality of trained object classifiers and preset audio data. The image processing module performs image detection to the image, and generates at least one segmented image, and then performs object detection and object recognition to generate a recognizing signal. The system processing module generates an audio signal to inform the user according to the recognizing signal.本發明係揭露一種具即時語音回報功能的導盲系統及其方法,利用二處理模組,分別執行影像處理及語音回報,以使該具即時語音回報功能的導盲系統即時回報路況。此系統包含:一影像感測模組、一記憶體、一影像處理模組、一系統處理模組及一音訊模組。影像感測模組感測一影像。記憶體儲存複數筆訓練樣本及複數筆預設音訊資訊。影像處理模組對影像進行一影像檢測,以產生至少一分割影像,再對至少一分割影像進行一物件偵測及一物件辨識,進而產生一辨識訊號。系統處理模組依據辨識訊號生一音訊訊號,以供一使用者聽取。10...具即時語音回報功能的導盲系統11...影像感測模組12...記憶體121...訓練樣本122...語音回報資訊13...影像處理模組131...影像檢測1311...分割影像132...物件偵測1321...物件特徵133...物件辨識1331...辨識訊號14...系統處理模組141...人機介面15...音訊模組151...語音音訊20...影像201...物件
