Abstract:
A network-based PLC(Packet Loss Concealment) method and an apparatus for the same are provided to maintain the quality of VoIP services regularly by increasing a PLC effect in a lost voice packet interval as a receiving unit executes PLC with loss-concealed voice packets having high correlation. A PLC apparatus(320) comprises a packet receiving part(321), a PLC processing part(322), a history buffer(323), and a packet transmitting part(324). The packet receiving part(321) receives VoIP packets, transmitted from a transmitting unit. The PLC processing part(322) judges whether packet loss has occurred in the VoIP packets received from the packet receiving part(321), and performs packet loss compensation for lost packets. The history buffer(323) stores the VoIP packets received from the PLC processing part(322). The packet transmitting part(324) forwards the VoIP packets, received from the PLC processing part(322), to the final receiving unit through a network.
Abstract translation:提供了一种基于网络的PLC(Packet Loss Concealment,分组丢包隐藏)方法及其装置,用于通过在丢失语音分组间隔中增加PLC效应来定期维护VoIP服务的质量,因为接收单元执行具有丢失隐藏语音分组的PLC 具有很高的相关性。 PLC装置(320)包括分组接收部分(321),PLC处理部分(322),历史缓冲器(323)和分组发送部分(324)。 分组接收部分(321)从发送单元接收VoIP分组。 PLC处理部(322)判断从分组接收部(321)接收到的VoIP分组中是否发生分组丢失,并对丢包进行分组丢失补偿。 历史缓冲器(323)存储从PLC处理部(322)接收的VoIP分组。 分组发送部(324)通过网络将从PLC处理部(322)接收的VoIP分组转发到最终的接收部。
Abstract:
PURPOSE: A method for restoring a transport error in an image and an apparatus thereof are provided to efficiently restore error generated when an image is transmitted under a communication environment, thereby restoring an original image. CONSTITUTION: A discriminating unit(20) obtains the standard deviation of pixel values about image blocks of each damaged block. The discriminating unit determines the complexity of a corresponding image block based on the obtained standard deviation. An error restoring unit(30) selects one of the entire pixel information and regional pixel information based on the complexity discriminated for each image block. The error restoring unit restores an error area included in a corresponding image block based on the selected information.
Abstract:
본 발명은 IP 네트워크 기반의 미디어 서비스에서 제공되는 영상의 품질을 측정하기 위한 장치 및 방법에 관한 것으로서, 더욱 상세하게는 전송영상과 원본영상을 수신하여 프레임을 추출하고, 추출된 프레임을 하위 블록으로 분할하여 휘도에 따른 랭크맵을 작성한 후, 프레임 간의 거리를 계산하여 원본영상과 매칭(matching)되는 전송영상의 특정구간을 검출하는 것을 특징으로 한다. 이렇게 검출된 전송영상의 특정구간과 원본영상을 이용하여 영상의 QoE 파라미터를 측정하면, 실시간으로 영상의 품질을 모니터링할 수 있기 때문에 보다 효율적으로 영상 품질을 측정할 수 있다. 전IP화(All-IP), IP 미디어(IP-media), QoE, 영상품질, 랭크맵
Abstract:
본 발명은 VoIP(Voice over Internet Protocol) 망에서 서비스 품질이 보장된 통화 음질 제공을 위한 통화 음질 측정 방법 및 시스템에 관한 것으로, 본 발명의 VoIP 망에서 통화 음질 측정 시스템은, IP를 이용한 패킷 전송을 통해 송수신 단 간에 음성 통화 서비스를 제공하는 음성통화망, 상기 송수신 단 간에 호가 설정된 상태에서 수신되는 통화음질측정 요구신호에 대해 참조음성파일을 생성하여 전송하고 상기 생성한 참조음성파일을 패킷화하여 시험음성데이터로 변환하여 전송하는 음질측정서버, 및 입력되는 명령에 따라 상기 통화음질측정 요구신호를 상기 음질측정서버로 전송하고 상기 음질측정서버로부터 전송된 참조음성파일을 기준으로 상기 시험음성데이터의 왜곡에 따라 발생하는 통화음질 저해요소와 기 저장된 통화음질 저해요소의 값을 비교하여 산출되는 편차 값에 기초해 상기 음성통화망의 장애에 의한 통화음질 저해요소를 제거하여 서비스 품질이 보장된 통화 음질 서비스를 제공하는 소프트폰 단말을 포함하며, 이에 따라, 보다 서비스 품질(QoS)이 좋은 음성통화품질을 제공할 수 있고 통화 품질 유지 보수를 위한 업무 효율을 증대시킬 수 있다. VoIP, PESQ, 통화 음질, 품질, 측정, QoS, 패킷, 음성
Abstract:
본 발명은 네트워크 기반의 패킷 손실 은닉 방법 및 이를 위한 장치에 관한 것으로서, 기존의 수신단에서만 패킷 손실 은닉 방법에 의하여 손실된 음성 패킷을 은닉하는 경우에 비해, 네트워크상에서 손실된 패킷과 상관도가 높은 음성 패킷을 가지고 수신단에서 패킷 손실 은닉을 미리 수행함으로써, 손실된 음성 패킷 구간에 대하여 패킷 손실 은닉 효과를 더욱 향상시킬 수 있어 VoIP 서비스의 품질을 일정하게 유지할 수 있도록 한 것이다. 패킷 손실 은닉, VoIP, 히스토리 버퍼, PLC
Abstract:
A method for measuring bi-directional interactive speech quality in a VoIP(Voice over Internet Protocol) network is provided to provide a high-quality VoIP service to a user by predicting accurate speech quality between ends. When measurement of speech quality is requested, a test VoIP packet for measuring speech quality is generated and transmitted to the other party(S41). A transmission side and a reception side receive the test VoIP packet and analyze it(S42). A damage value(Ie) caused by a packet loss is calculated through a PESQ(Perceptual Evaluation of Speech Quality) algorithm(S43). A delay value(Id) in accordance with an end-to-end packet delay is calculated based on an actual VoIP network(S44). A jitter value(Ij) in accordance with irregularity of the end-to-end packet delay is calculated based on the actual VoIP network(S45). A speech quality measurement value between a transmitter and a receiver is calculated by applying the damage value(Ie), the delay value(Id) and the jitter value(Ij) to a new E-Model(S46).
Abstract:
A method and a system for measuring voice quality with QoS(Quality of Service) assurance in a VoIP(Voice over Internet Protocol) network are provided to offer voice quality having good QoS by calculating values related to parameters detrimental to voice quality when a voice quality measurement request is made, judging the fault state of a VoIP network connected between users, and removing fault factors. A system for measuring voice quality with QoS assurance comprises a VoIP network(380), a voice quality measurement server(500), and soft phone terminals(100,200). The VoIP network(380) provides a voice call service between transmitting and receiving units through packet transmission using an IP. The voice quality measurement server(500) creates and transmits a reference voice file for a voice quality measurement request received in the middle of telephone conversation between the transmitting and receiving units, packetizes the created reference voice file, converts it into test voice data, and transmits the converted data. Either of the soft phone terminals(100,200) transmits the voice quality measurement request to the voice quality measurement server(500) according to an inputted command, removes voice quality deterioration parameters, due to a fault of the VoIP network(380) generated according to distortion of the test voice data, on the basis of the reference voice file, and provides a voice quality service with QoS assurance.
Abstract:
A noise removal apparatus for a VoIP(Voice over Internet Protocol) service and a method therefor are provided to remove a noise signal in consideration of an SNR(Signal-to-Noise Ratio) weight value. A power spectrum estimator(110) estimates a voice power spectrum and a noise power spectrum from a noise voice signal. A Wiener filter designing unit(120) calculates an SNR from the voice power spectrum and the noise power spectrum, acquires an SNR weight value corresponding to the calculated SNR, and designs a Wiener filter function which reflects the SNR weight value. A Wiener filter unit(130) removes a noise signal included in the noise voice signal by using the Wiener filter function.
Abstract:
A speech noise canceling apparatus and method for guaranteeing QoS(Quality of Service) of a VoIP service, and a VoIP(Voice over Internet Protocol) terminal employing the same are provided to measure noise generated between VoIP soft phones by using a PESQ(Perceptual Evaluation of Speech Quality) speech quality measurement algorithm and a voice data code book and cancel its noise. An input unit(122) receives a cancel signal of noise generated during voice call communication from the user. A file/data generating unit(142) generates a reference voice file corresponding to actual voice data in a VoIP network, packetizes the reference voice file, and converts the packetized reference voice file into test voice data. A PESQ measurement unit(140) extracts a PESQ-MOS(Mean Opinion Score) value by applying the reference voice file and the test voice data to a pre-set PESQ speech quality measurement algorithm. A parameter calculating unit(150) calculates a speech quality disincentive value corresponding to the generated noise by using the extracted PESQ-MOS value. A deviation value calculating unit(160) calculates the deviation between the calculated speech quality disincentive value and a pre-set speech quality disincentive value. A network management unit(170) determines an error state of the VoIP network based on the calculated deviation value and removes the speech quality disincentive.
Abstract:
PURPOSE: A method for measuring block distortion based on a reduced reference scheme is provided to support quick processing speed using a reduced reference scheme for measuring the quality of a received processing image. CONSTITUTION: An edge image of an original image is extracted. A block edge is extracted(411-414). A weighted value of the extracted block edge is determined. A block distortion value of the original image is calculated using the calculated weighted value(415). The received image edge image is extracted(422). The gradient size and direction of the edge image are calculated. A block edge is extracted(424).