1.添加Vision Kit依赖

在项目设置中点击"General"选项卡，然后在"Frameworks, Libraries, and Embedded Content"（框架、库和嵌入内容）部分，点击"+"按钮。搜索并选择"Vision.framework"。

2.代码实现

import Vision

func recognizeText (image: CGImage){ 

    let request = VNRecognizeTextRequest { (request , error) in 
        if let results = request.results as ?[VNRecognizedTextObservation]{
        //处理识别结果 
        let recognizedStrings = results.compactMap { observation in 
            return observation.topCandidates(1).first?.string 
        } 
        print ("识别结果：\(recognizedStrings)") 
        self.result.text = recognizedStrings.joined (separator:"\n")
        }
    }

    //设置最小置信度来过滤结果 
    request.minimumTextHeight =1 
    //精确识别，fast则为快速识别
    request.recognitionLevel =.accurate

    //创建图片请求处理器 
    let requestHandler = VNImageRequestHandler(cgImage: image , options :[:])

    //执行请求 
    do { 
        try requestHandler.perform ([request])
    } catch {
        print ("文字识别失败：\(error)")
    }
}

3.识别结果

速度很快，但亲测识别准确率有限。部分识别文字错误。

提高中文识别成功率，可以设置

request.recognitionLanguages = ["zh-Hans"]