Question

我有一个coreml模型，该模型运行后将返回带有1个“ MultiArray：Double 10 x IMG_SIZE x IMG_SIZE数组”的VNCoreMLFeatureValueObservation对象

如何将其转换为10个UIImage，每个UIImage具有IMG_SIZE x IMG_SIZE尺寸，并且它们的值是灰度的？

Answer 1

窥探了一下之后，我发现我必须添加以下辅助函数：

https://github.com/hollance/CoreMLHelpers到我的Xcode项目。并且来自MultiArray初始化问题：https://stackoverflow.com/a/44462908/403403

然后我将这个解决方案拼凑在一起：

 let request = VNCoreMLRequest(model: model) { (request, error) in
            guard let results = request.results as? [VNCoreMLFeatureValueObservation] else {
                fatalError("Model failed to process image")
            }

            let obs : VNCoreMLFeatureValueObservation = (results.first)!
            let m: MLMultiArray = obs.featureValue.multiArrayValue!
            var mArrays = [MLMultiArray]()


            for i in 1..<10 {
                let start = i*(IMG_SIZE*IMG_SIZE) 
                guard let tmp : MLMultiArray = try? MLMultiArray(shape:[768,768], dataType:MLMultiArrayDataType.double) else {
                    fatalError("Unexpected runtime error. MLMultiArray")
                }
                for n in 0..<(IMG_SIZE*IMG_SIZE) {
                    tmp[n] = m[start+n]
                }
                mArrays.append(tmp)
            }

            guard let mlMultiArray : MLMultiArray = try? MLMultiArray(shape:[768,768], dataType:MLMultiArrayDataType.double) else {
                fatalError("Unexpected runtime error. MLMultiArray")
            }

            for n in 0..<(IMG_SIZE*IMG_SIZE) {
                mlMultiArray[n] = m[n]
            }

            let imagePredict : UIImage = mlMultiArray.image(offset: 0, scale: 255)!

            self.imagePred1.image = imagePredict
            self.imagePred2.image = mArrays[0].image(offset: 0, scale: 255)!
            self.imagePred3.image = mArrays[1].image(offset: 0, scale: 255)!
            self.imagePred4.image = mArrays[2].image(offset: 0, scale: 255)!
            self.imagePred5.image = mArrays[3].image(offset: 0, scale: 255)!
            self.imagePred6.image = mArrays[4].image(offset: 0, scale: 255)!
            self.imagePred7.image = mArrays[5].image(offset: 0, scale: 255)!
            self.imagePred8.image = mArrays[6].image(offset: 0, scale: 255)!
            self.imagePred9.image = mArrays[7].image(offset: 0, scale: 255)!



        }

希望有一种更清洁的方法，但是现在可以使用

将VNCoreMLFeatureValueObservation结果（3D双数组）转换为多个UIImage

1 个答案: