安卓 java。lang.NullPointerException:尝试在空对象引用上调用接口方法
下面的代码抛出一个NullPointerException,我不知道为什么。如果有人能给我指出错误,我将不胜感激
来自MainActivity的代码,错误在第4行:
private void init(){
this.createClassifier();
this.takePhoto();
}
private void createClassifier(){
try {
classifier = ImageClassifierFactory.create(
getAssets(),
Constants.GRAPH_FILE_PATH,
Constants.LABELS_FILE_PATH,
Constants.IMAGE_SIZE,
Constants.GRAPH_INPUT_NAME,
Constants.GRAPH_OUTPUT_NAME);
}catch (IOException e) {
Log.e("MainActivity", e.getMessage());
}
}
private Classifier classifier;
...
private final void classifyAndShowResult(final Bitmap croppedBitmap){
runInBackground((new Runnable(){
public final void run(){
Result result = classifier.recognizeImage(croppedBitmap);
showResult(result);
}
}));
}
有不同的方法调用init(),所以我不明白为什么分类器没有初始化
结果类:
public class Result {
private String result;
private float confidence;
public Result(String result, float confidence) {
this.result = result;
this.confidence = confidence;
}
public String getResult() {
return result;
}
public void setResult(String result) {
this.result = result;
}
public float getConfidence() {
return confidence;
}
public void setConfidence(float confidence) {
this.confidence = confidence;
}
}
分类器接口:
public interface Classifier {
Result recognizeImage(Bitmap bitmap);
以及启动的识别图像:
public Result recognizeImage(Bitmap bitmap){
preprocessImageToNormalizedFloats(bitmap);
classifyImageToOutputs();
PriorityQueue outputQueue = new PriorityQueue(getResults());
Object queue = outputQueue.poll();
return (Result)queue;
}
错误代码是:java。lang.NullPointerException:尝试调用接口方法的分类器。结果分类器。分类器。在空对象引用上识别图像(安卓.graphics.Bitmap)”
ImageClassifier构造函数:
public ImageClassifier(String inputName, String outputName, long imageSize, List labels, int[] imageBitmapPixels, float[] imageNormalizedPixels, float[] results, TensorFlowInferenceInterface tensorFlowInference) {
this.inputName = inputName;
this.outputName = outputName;
this.imageSize = imageSize;
this.labels = labels;
this.imageBitmapPixels = imageBitmapPixels;
this.imageNormalizedPixels = imageNormalizedPixels;
this.results = results;
this.tensorFlowInference = tensorFlowInference;
}
ImageClassifierFactory类:
public class ImageClassifierFactory {
public final static Classifier create(AssetManager assetManager, String graphFilePath, String labelsFilePath, int imageSize, String inputName, String outputName) throws IOException {
List labels = getLabels(assetManager, labelsFilePath);
ImageClassifier im = new ImageClassifier(inputName, outputName, (long)imageSize, labels, new int[imageSize * imageSize], new float[imageSize * imageSize * COLOR_CHANNELS], new float[labels.size()], new TensorFlowInferenceInterface(assetManager, graphFilePath));
return im;
}
}
# 1 楼答案
无法创建接口分类器的分类器实例。 您需要将分类器更改为类并创建该类的分类器实例,或者创建实现分类器接口的另一个类的实例
# 2 楼答案
声明
classifyAndShowResult
-方法的类的属性中必须有classifier
。在调用该方法时,该变量是null
。要对此进行调试,您应该在“调试”模式下启动应用程序(假设您在eclipse中编程),然后读取NullPointerException
的堆栈以找出classifier
尚未启动的原因@Edit: 在看到为分类器截取的代码后,您需要执行以下操作:
。。。假设
classifier
在初始化时不需要参数