这里是一个工作示例，就像我之前说的，这可能是一个时间问题，因为你在body中工作。
但另一个需要考虑的是，VNRecognizeTextRequest需要JPEG，因为它没有alpha/透明度。

import SwiftUI
import Vision
import VisionKit
struct TextRecognitionView: View {
    let model: TextRecognitionModel = .init()
    @State private var image: UIImage?
    @State private var recognizedText: String? = nil
    var body: some View {
        VStack(spacing: 15){
            VStack {
                switch image { //Unwrap the Image
                case .none:
                    Text("You haven't taken any pictures yet!")
                case .some(let image):
                    Image(uiImage: image)
                        .resizable()
                        .scaledToFit()
                    switch recognizedText { //Unwrap the text
                    case .none:
                        ProgressView() //Show this while recognizing
                    case .some(let text):
                        Text(text) //Show text
                    }
                }
                
                Button("set random text image") {
                    self.image = Text("Random text \((0...100).randomElement()!)")
                        .frame(width: 100, height: 100)
                        .snapshot().validJPEG() // Vision requires JPEG image likely because of transparency/alpha
                    self.recognizedText = nil //Clear the text
                    Task {
                        do {
                            self.recognizedText = try await model.performRequest(image: image!)
                        } catch {
                            self.recognizedText = error.localizedDescription //Show an error to the user
                            print(error)
                        }
                    }
                }
            }
        }
    }
}

struct TextRecognitionModel {
    /// async await version of a VNRecognizeTextRequest + VNImageRequestHandler
    func performRequest(image: UIImage) async throws -> String {
        guard let cgImage: CGImage = image.cgImage  else {
            throw RequestErrors.unableToRetrieveImage
        }
        let requestHandler: VNImageRequestHandler = VNImageRequestHandler(cgImage: cgImage)
        
        return try await withCheckedThrowingContinuation({ continuation in
            let request = VNRecognizeTextRequest(completionHandler: { request, error in
                if let error {
                    continuation.resume(throwing: error)
                } else {
                    let results = request.results as? [VNRecognizedTextObservation] ?? []
                    let recognizedStrings: Array<String> = results.compactMap({result in result.topCandidates(1).first!.string})
                    
                    continuation.resume(returning: recognizedStrings.joined())
                }
            })
            do {
                try requestHandler.perform([request])
            } catch {
                continuation.resume(throwing: error)
            }
        })
    }
}

enum RequestErrors: LocalizedError {
    case unableToRetrieveImage
}

struct TextRecognitionView_Previews: PreviewProvider {
    static var previews: some View {
        TextRecognitionView()
    }
}

extension View {
    func snapshot() -> UIImage {
        let controller = UIHostingController(rootView: self)
        let view = controller.view
        
        let targetSize = controller.view.intrinsicContentSize
        view?.bounds = CGRect(origin: .zero, size: targetSize)
        view?.backgroundColor = .clear
        
        let renderer = UIGraphicsImageRenderer(size: targetSize)
        
        return renderer.image { _ in
            view?.drawHierarchy(in: controller.view.bounds, afterScreenUpdates: true)
        }
    }
}

extension UIImage {
    func validJPEG() -> UIImage {
        guard let data = self.jpegData(compressionQuality: 1) else {
            return .init()
        }
        
        guard let jpegImage = UIImage(data: data) else {
            return .init()
        }
        return jpegImage
    }
}