kotlin 中 tensorflow lite 边界框的绘制未与对象对齐

Question

我正在创建一个具有对象检测和捕获图像的应用程序。我按照 https://www.tensorflow.org/lite/android/tutorials/object_detection 和 https://github.com/tensorflow/examples/tree/master/lite/examples/object_detection/android 作为示例代码。

class PlotBoundingBox(
    context: Context,
    private val detectedObject: List<Detection>,
    private val imageHeight: Int,
    private val imageWidth: Int
) : View(context) {

    private var boundaryPaint = Paint()
    var scaleFactor = 1f

    init {
        init()
    }

    private fun init() {
        boundaryPaint.color = Red.hashCode()
        boundaryPaint.strokeWidth = 5f
        boundaryPaint.style = Paint.Style.STROKE
    }

    @SuppressLint("DrawAllocation")
    override fun onDraw(canvas: Canvas) {
        super.onDraw(canvas)

        
        scaleFactor = max(width * 1f / imageWidth, height * 1f / imageHeight)

        for (result in detectedObject) {
            val boundingBox = result.boundingBox

            val top = boundingBox.top * scaleFactor
            val bottom = boundingBox.bottom * scaleFactor
            val left = boundingBox.left * scaleFactor
            val right = boundingBox.right * scaleFactor

            // Draw bounding box around detected objects
            val drawableRect = RectF(left, top, right, bottom)

            canvas.drawRect(drawableRect, boundaryPaint)
        }
    }
}

我能够正确绘制边界框，但是当我捕获图像时，图像与预览视图中的显示不一样（与预览视图中的显示相比，捕获的图像被调整到左侧）。所以我决定将比例设置为 FILL_CENTER，但与使用 FILL_START 不同，边界框没有正确绘制。以下是我对我的实现代码的参考。我对边界框绘图的参考在这里https://github.com/tensorflow/examples/blob/master/lite/examples/object_detection/android/app/src/main/java/org/tensorflow/lite/examples /objectdetection/OverlayView.kt.