Who helps a letter find its way when a machine cannot read its address?
机器读不懂的信，谁来帮它找路？
机器读不懂一封信上的地址，这封信就走不下去了吗？
If a machine cannot read the address on a letter, does that letter have to stop there?
在美国邮政的处理流程里，有一群人专门帮助解决这个问题。
In the U.S. Postal Service's processing system, a group of people helps solve this problem.
有意思的是，他们收到的是地址的图像，而不是送到眼前的那封信。
Interestingly, they receive an image of the address rather than the physical letter brought to them.
这些人工作的地方，是盐湖城的远程编码中心。
These people work at the Remote Encoding Center in Salt Lake City.
当自动设备读不出地址时，处理地点会拍下信件的图像，发送给这个中心。
When automated equipment cannot read an address, the processing facility takes an image of the letter and sends it to this center.
工作人员看着图像，把能辨认出的地址信息输入电脑。
Workers look at the image and enter the address information they can identify into a computer.
他们要处理的是信件外面的地址，不是信里写了什么。
They deal with the address on the outside of the letter, not what is written inside it.
输入的信息会被拿来和有效的地址对照，找出对应的是哪个地址。
The entered information is compared with valid addresses to find the corresponding address.
然后，数据传回原来的处理地点，用来在信件上印出条码。
The data then returns to the original processing facility to print a barcode on the letter.
条码上有一条条线，机器读这些线，就能获得信息。
A barcode has a series of lines; machines read these lines to obtain information.
设备能读条码，信件就可以继续进入自动处理流程。
Once the equipment can read the barcode, the letter can continue in the automated processing system.
这并不表示所有模糊的地址都一定能认出来。
This does not mean that every unclear address can definitely be identified.
不过，图像可以先去找人帮忙，信件不用为这一步搬到那个中心。
The image can go to people for help, while the letter does not have to travel to that center for this step.
自动分信的背后，还藏着人把难读的字变成机器能读的信息这一步。
Behind automated mail sorting is also a step in which people turn hard-to-read writing into information machines can read.
