Penang to equip CCTV cameras with facial recognition in bid to reduce crime

Penang to equip CCTV cameras with facial recognition in bid to reduce crime

Penang police facial recognition
Penang chief minister Chow Kon Yeow at the launch of a facial recognition system capable of detecting faces of criminals through CCTV cameras (Photo: Bernama)

GEORGE TOWN: Penang has launched a facial recognition system capable of detecting the faces of known criminals through closed-circuit television (CCTV) cameras.

It is the first state in Malaysia to launch such a system, which relies on artificial intelligence technology developed by IBM. 

Speaking at the launch event on Wednesday (Jan 2), chief minister Chow Kon Yeow said the technology would enhance the capabilities of the existing 767 CCTV cameras installed by the Penang City Council. 

The system will be operated from the council's CCTV control room, as well as the Penang police headquarters, he said. 

"The monitoring via CCTVs is an initiative by the Penang state government to reduce crime, especially street crimes, in an effort to maintain the safety and well-being of the people," said Chow.

Since 2008, the installation of CCTV cameras has cost the state RM46.2 million (US$11.16 million), the minister said. The facial recognition system as well as the installation of another 150 CCTV cameras would require RM12 million. 

The project is expected to be completed by May.


State police chief A Thaiveegan, who was also present at the launch, said the facial recognition technology would help increase the police's efficiency.

He said the police would upload the image of a criminal or wanted person to the system. If the system manages to track similar faces through any of the CCTV cameras, the authorities would immediately be alerted. 

"For example, in snatch theft cases, the police may not be able to catch the criminals during the time of the incident, but through the technology, we can clearly identify the suspect and it will be easier track him or her down," he added.

Thaiveegan said that the police would manually track the movement of the suspects before arresting them.

“We do not go straight to the person and arrest them. We have to check their identification cards and crosscheck with our database on the wanted list before taking action.

“The cameras usually trigger when it reaches 80 to 90 per cent accuracy and even that depends on the camera quality,” the police chief was quoted as saying by the New Straits Times.

Facial recognition systems have been implemented in other parts of the world in support of security operations.

The Chinese government, for instance, is working to combine its 170 million security cameras with facial recognition and artificial intelligence technology.

READ: ICA to introduce multi-biometrics screening, facial recognition at checkpoints from April 2019

Source: Bernama/CNA/aw(hs)