為什么閉路電視畫面仍然如此垃圾?

為什么閉路電視畫面仍然如此垃圾?

The number of CCTV cameras in the UK has been estimated at up to six million, most of which are installed in private homes and businesses. All CCTV cameras – whether the older, analogue type or the more modern, digital ones – have to make trade-offs between picture quality, storage and cost.

據(jù)估計(jì)，英國有多達(dá)600萬個(gè)閉路電視攝像頭，其中大部分安裝在私人住宅和企業(yè)里。所有的閉路電視攝像機(jī)——無論是老式的模擬攝像機(jī)還是更現(xiàn)代的數(shù)字?jǐn)z像機(jī)——都必須在畫面質(zhì)量、存儲(chǔ)空間和成本之間做出權(quán)衡。

There are around 13,000 cameras in London Underground stations alone, and the footage from these is kept for two weeks. Upgrading all of these to record at 4K quality instead of 1080p would require an extra 300 million gigabytes of storage, not to mention the cost of the cameras themselves.

僅在倫敦地鐵站就有大約13000個(gè)攝像頭，這些攝像頭拍攝的視頻被保存了兩周。將所有這些設(shè)備升級(jí)到4K質(zhì)量，而不是1080p，將需要額外3億gb的存儲(chǔ)空間，更不用說攝像頭本身的成本了。

CCTV cameras also generally use wide-angle lenses to cover the broadest possible area. Digitally zooming in on a single person in the distance inevitably results in a blurrier image. Nevertheless, image quality is increasing all the time, and image processing software and machine learning algorithms can now recognise faces even from low-resolution video.

閉路電視攝像機(jī)通常也使用廣角鏡頭來覆蓋盡可能廣泛的區(qū)域。數(shù)字放大一個(gè)人在遠(yuǎn)處不可避免地導(dǎo)致一個(gè)模糊的形象。然而，圖像質(zhì)量一直在提高，圖像處理軟件和機(jī)器學(xué)習(xí)算法現(xiàn)在甚至可以從低分辨率視頻中識(shí)別人臉。