Android Question Help implementing ML Kit Object Detection

Lucas Siqueira · Jul 11, 2025

Hello everyone, Erel did a wonderful job implementing several classes with ML Kit, but I've been trying to implement Object Detection for 3 days, but I haven't been able to, can anyone do it?

documentation:

Detect and track objects with ML Kit on Android | Google for Developers

developers.google.com

Erel has already implemented:

[class] ML Kit Subject Segmentation

https://developers.google.com/ml-kit/vision/subject-segmentation Input: image Output: image with main subject See attached project. Don't miss the #AdditionalJars in the main module and the manifest editor snippets. Usage: Private Sub Button1_Click Dim cc As ContentChooser...

www.b4x.com

[B4X] SelfieSegmentation with ML Kit

Image source: https://en.wikipedia.org/wiki/Selfie#/media/File:TWC_Hokitika_Gorge_•_Stewart_Nimmo_•_MRD_1.jpg A B4A + B4i selfie segmentation feature based on Google ML Kit: https://developers.google.com/ml-kit/vision/selfie-segmentation The result is a mask image. In the example it is...

www.b4x.com

Digital Ink Recognition based on ML Kit

This is a B4A class that wraps ML Kit digital ink recognition. More information here: https://developers.google.com/ml-kit/vision/digital-ink-recognition This feature has a large set of dependencies. I recommend starting with a simpler feature such as this one...

www.b4x.com

[B4X] TextRecognition based on MLKit

This is a B4A + B4i solution. It recognizes text in images based on Google ML Kit. Dependencies: - B4i (local Mac only): https://www.b4x.com/b4i/files/MLKit.zip - B4A: open B4A Sdk Manager - search for mlkit and install all items. Also search and install odml. Don't miss: - B4i...

www.b4x.com

TILogistic · Jul 11, 2025

?

Detecção e rastreamento de objetos | ML Kit | Google for Developers

developers.google.com

read:
Como usar um modelo personalizado do TensorFlow Lite

Lucas Siqueira · Jul 11, 2025

I couldn't implement it in b4x, please help me to implement it

Object detection and tracking | ML Kit | Google for Developers

developers.google.com

jkhazraji · Jul 12, 2025

Lucas Siqueira said:
I couldn't implement it in b4x, please help me to implement it

Object detection and tracking | ML Kit | Google for Developers

developers.google.com

Before I came across this post, I was also digging in the same subject, but it seems that b4a does not lend itself to ML Kit easily although it deals with android for if you want to do the same thing in Android Studio, ML Kit detect object is easily integrated with 'implementation' statement. Therefore, depending on the examples from Google and Github I came out with a b4a code (manly inline Java) but still it has one stubborn error failure to initialize ML Kit:
Exception during ML Kit Initialization :Attempt to invoke virtual method 'com.google.mlkit.vision.common.internal.MultiFlavorDetectorCreator$MultiFlavorDetector com.google.mlkit.vision.common.internal.MultiFlavorDetectorCreator.create(com.google.mlkit.vision.common.internal.MultiFlavorDetectorCreator$DetectorOptions)' on a null object reference.
If you or the forum experts find a solution, sharing it will be of benefit.

B4X:

#Region  Project Attributes
    #ApplicationLabel: B4A ML Kit Detection
    #VersionCode: 1
    #VersionName: 1.0
    #SupportedOrientations: unspecified
    #CanInstallToExternalStorage: False
    
#End Region

#Region  Activity Attributes
    #FullScreen: False
    #IncludeTitle: True
#End Region

#AdditionalJar: com.google.mlkit:object-detection
#AdditionalJar: com.google.mlkit:vision-common
#AdditionalJar: com.google.android.gms:play-services-base
#AdditionalJar: com.google.android.gms:play-services-basement
#AdditionalJar: com.google.android.gms:play-services-tasks
#AdditionalJar: androidx.annotation:annotation
#AdditionalJar: kotlin-stdlib-1.6.10

Sub Process_Globals
    Private xui As XUI
    Private isMLKitInitialized As Boolean = False
End Sub

Sub Globals
    Private iv As ImageView
    Private btnDetect As Button
    Private lblResults As Label
End Sub

Sub Activity_Create(FirstTime As Boolean)
    Activity.LoadLayout("Layout")
    
    ' Initialize views
    iv.Initialize("iv")
    btnDetect.Initialize("btnDetect")
    lblResults.Initialize("lblResults")
    
    Activity.AddView(iv, 0, 0, 100%x, 60%y)
    Activity.AddView(btnDetect, 0, 65%y, 100%x, 15%y)
    Activity.AddView(lblResults, 0, 85%y, 100%x, 15%y)
    
    btnDetect.Text = "Detect Objects"
    btnDetect.Enabled = False
    lblResults.Text = "Initializing ML Kit..."
    
    ' Copy test image
    If Not(File.Exists(File.DirInternal,"home.jpg")) Then
        File.Copy(File.DirAssets,"home.jpg",File.DirInternal,"home.jpg")
    End If
    
    ' Load test image
    Dim bmp As Bitmap = LoadBitmap(File.DirInternal, "test.png")
    iv.Bitmap = bmp
    
    ' Wait for activity to be fully loaded, then initialize ML Kit
    Wait For Activity_Resume
    Sleep(1000) ' Give the activity time to fully load
    InitializeMLKit
End Sub

Sub InitializeMLKit
    Try
        Dim jo As JavaObject
        jo.InitializeContext
        jo.RunMethod("initializeMLKit", Array(GetBA))
    Catch
        Log("Error initializing ML Kit: " & LastException.Message)
        lblResults.Text = "ML Kit initialization error: " & LastException.Message
    End Try
End Sub

Sub btnDetect_Click
    If Not(isMLKitInitialized) Then
        lblResults.Text = "ML Kit not initialized yet"
        Return
    End If
    
    If iv.Bitmap <> Null Then
        lblResults.Text = "Detecting objects..."
        DetectObjects(iv.Bitmap)
    Else
        lblResults.Text = "No image loaded"
    End If
End Sub

Sub DetectObjects(bmp As Bitmap)
    Try
        Dim jo As JavaObject = Me
        jo.InitializeContext
        jo.RunMethod("detectObjects", Array(bmp, GetBA))
    Catch
        Log("Error in DetectObjects: " & LastException.Message)
        lblResults.Text = "Detection error: " & LastException.Message
    End Try
End Sub

Sub GetBA As JavaObject
    Dim jo As JavaObject
    Dim cls As String = Me
    cls = cls.SubString("class ".Length)
    jo.InitializeStatic(cls)
    Return jo.GetFieldJO("processBA")
End Sub

#If JAVA
import com.google.mlkit.common.MlKit;
import com.google.mlkit.vision.common.InputImage;
import com.google.mlkit.vision.objects.DetectedObject;
import com.google.mlkit.vision.objects.ObjectDetection;
import com.google.mlkit.vision.objects.ObjectDetector;
import com.google.mlkit.vision.objects.defaults.ObjectDetectorOptions;
import com.google.android.gms.tasks.OnSuccessListener;
import com.google.android.gms.tasks.OnFailureListener;
import com.google.android.gms.tasks.Task;
import android.graphics.Bitmap;
import java.util.List;
import anywheresoftware.b4a.BA;
import android.content.Context;
import android.util.Log;

private static ObjectDetector objectDetector = null;

public void initializeMLKit(BA ba) {
    try {
        Context context = ba.context;
        BA.Log("Starting ML Kit initialization...");
        
        MlKit.initialize(context);
        BA.Log("ML Kit initialized");
        
        ObjectDetectorOptions options = new ObjectDetectorOptions.Builder()
            .setDetectorMode(ObjectDetectorOptions.SINGLE_IMAGE_MODE)
            .build();
        
        BA.Log("Creating object detector...");
        
        objectDetector = ObjectDetection.getClient(options);
        
        if (objectDetector != null) {
            BA.Log("Object detector created successfully");
            ba.raiseEvent(this, "initialization_complete");
        } else {
            BA.Log("Failed to create object detector - returned null");
            ba.raiseEvent(this, "initialization_failed", "ObjectDetection.getClient returned null");
        }
        
    } catch (Exception e) {
        BA.Log("Exception during ML Kit initialization: " + e.getMessage());
        e.printStackTrace();
        ba.setLastException(e);
        ba.raiseEvent(this, "initialization_failed", e.getMessage());
    }
}

public void detectObjects(Bitmap bitmap, BA ba) {
    try {
        BA.Log("Starting object detection...");
        
        if (objectDetector == null) {
            BA.Log("Object detector is null");
            ba.raiseEvent(this, "detection_error", "Object detector not initialized");
            return;
        }
        
        if (bitmap == null) {
            BA.Log("Bitmap is null");
            ba.raiseEvent(this, "detection_error", "Bitmap is null");
            return;
        }
        
        if (bitmap.isRecycled()) {
            BA.Log("Bitmap is recycled");
            ba.raiseEvent(this, "detection_error", "Bitmap is recycled");
            return;
        }
        
        BA.Log("Bitmap size: " + bitmap.getWidth() + "x" + bitmap.getHeight());
        
        InputImage image = InputImage.fromBitmap(bitmap, 0);
        BA.Log("InputImage created");
        
        Task<List<DetectedObject>> task = objectDetector.process(image);
        
        task.addOnSuccessListener(new OnSuccessListener<List<DetectedObject>>() {
            @Override
            public void onSuccess(List<DetectedObject> detectedObjects) {
                BA.Log("Detection successful, found " + detectedObjects.size() + " objects");
                
                StringBuilder results = new StringBuilder();
                
                if (detectedObjects.isEmpty()) {
                    results.append("No objects detected");
                } else {
                    results.append("Found ").append(detectedObjects.size()).append(" objects:\n\n");
                    
                    int objectCount = 1;
                    for (DetectedObject detectedObject : detectedObjects) {
                        results.append("Object ").append(objectCount).append(":\n");
                        
                        results.append("  Position: ")
                              .append(detectedObject.getBoundingBox().toString())
                              .append("\n");
                        
                        List<DetectedObject.Label> labels = detectedObject.getLabels();
                        if (!labels.isEmpty()) {
                            results.append("  Classifications:\n");
                            for (DetectedObject.Label label : labels) {
                                results.append("    ")
                                      .append(label.getText())
                                      .append(" (")
                                      .append(String.format("%.1f", label.getConfidence() * 100))
                                      .append("%)\n");
                            }
                        } else {
                            results.append("  No classification available\n");
                        }
                        
                        results.append("\n");
                        objectCount++;
                    }
                }
                
                ba.raiseEvent(this, "detection_success", results.toString());
            }
        });
        
        task.addOnFailureListener(new OnFailureListener() {
            @Override
            public void onFailure(Exception e) {
                BA.Log("Detection failed: " + e.getMessage());
                e.printStackTrace();
                ba.raiseEvent(this, "detection_failed", e.getMessage());
            }
        });
        
    } catch (Exception e) {
        BA.Log("Exception during object detection: " + e.getMessage());
        e.printStackTrace();
        ba.raiseEvent(this, "detection_error", e.toString());
    }
}
#End If

Sub initialization_complete
    isMLKitInitialized = True
    btnDetect.Enabled = True
    lblResults.Text = "ML Kit initialized successfully! Ready to detect objects."
    Log("ML Kit initialized successfully")
End Sub

Sub initialization_failed(Error As String)
    isMLKitInitialized = False
    btnDetect.Enabled = False
    lblResults.Text = "Initialization failed: " & Error
    Log("Initialization failed: " & Error)
End Sub

Sub detection_success(Results As String)
    lblResults.Text = Results
    Log("Detection results: " & Results)
End Sub

Sub detection_failed(Error As String)
    lblResults.Text = "Detection failed: " & Error
    Log("Detection failed: " & Error)
End Sub

Sub detection_error(Error As String)
    lblResults.Text = "Error: " & Error
    Log("Error: " & Error)
End Sub

Sub Activity_Resume
End Sub

Sub Activity_Pause (UserClosed As Boolean)
End Sub

Johan Schoeman · Jul 13, 2025

jkhazraji said:

Before I came across this post, I was also digging in the same subject, but it seems that b4a does not lend itself to ML Kit easily although it deals with android for if you want to do the same thing in Android Studio, ML Kit detect object is easily integrated with 'implementation' statement. Therefore, depending on the examples from Google and Github I came out with a b4a code (manly inline Java) but still it has one stubborn error failure to initialize ML Kit:
Exception during ML Kit Initialization :Attempt to invoke virtual method 'com.google.mlkit.vision.common.internal.MultiFlavorDetectorCreator$MultiFlavorDetector com.google.mlkit.vision.common.internal.MultiFlavorDetectorCreator.create(com.google.mlkit.vision.common.internal.MultiFlavorDetectorCreator$DetectorOptions)' on a null object reference.
If you or the forum experts find a solution, sharing it will be of benefit.

B4X:

#Region  Project Attributes
    #ApplicationLabel: B4A ML Kit Detection
    #VersionCode: 1
    #VersionName: 1.0
    #SupportedOrientations: unspecified
    #CanInstallToExternalStorage: False
   
#End Region

#Region  Activity Attributes
    #FullScreen: False
    #IncludeTitle: True
#End Region

#AdditionalJar: com.google.mlkit:object-detection
#AdditionalJar: com.google.mlkit:vision-common
#AdditionalJar: com.google.android.gms:play-services-base
#AdditionalJar: com.google.android.gms:play-services-basement
#AdditionalJar: com.google.android.gms:play-services-tasks
#AdditionalJar: androidx.annotation:annotation
#AdditionalJar: kotlin-stdlib-1.6.10

Sub Process_Globals
    Private xui As XUI
    Private isMLKitInitialized As Boolean = False
End Sub

Sub Globals
    Private iv As ImageView
    Private btnDetect As Button
    Private lblResults As Label
End Sub

Sub Activity_Create(FirstTime As Boolean)
    Activity.LoadLayout("Layout")
   
    ' Initialize views
    iv.Initialize("iv")
    btnDetect.Initialize("btnDetect")
    lblResults.Initialize("lblResults")
   
    Activity.AddView(iv, 0, 0, 100%x, 60%y)
    Activity.AddView(btnDetect, 0, 65%y, 100%x, 15%y)
    Activity.AddView(lblResults, 0, 85%y, 100%x, 15%y)
   
    btnDetect.Text = "Detect Objects"
    btnDetect.Enabled = False
    lblResults.Text = "Initializing ML Kit..."
   
    ' Copy test image
    If Not(File.Exists(File.DirInternal,"home.jpg")) Then
        File.Copy(File.DirAssets,"home.jpg",File.DirInternal,"home.jpg")
    End If
   
    ' Load test image
    Dim bmp As Bitmap = LoadBitmap(File.DirInternal, "test.png")
    iv.Bitmap = bmp
   
    ' Wait for activity to be fully loaded, then initialize ML Kit
    Wait For Activity_Resume
    Sleep(1000) ' Give the activity time to fully load
    InitializeMLKit
End Sub

Sub InitializeMLKit
    Try
        Dim jo As JavaObject
        jo.InitializeContext
        jo.RunMethod("initializeMLKit", Array(GetBA))
    Catch
        Log("Error initializing ML Kit: " & LastException.Message)
        lblResults.Text = "ML Kit initialization error: " & LastException.Message
    End Try
End Sub

Sub btnDetect_Click
    If Not(isMLKitInitialized) Then
        lblResults.Text = "ML Kit not initialized yet"
        Return
    End If
   
    If iv.Bitmap <> Null Then
        lblResults.Text = "Detecting objects..."
        DetectObjects(iv.Bitmap)
    Else
        lblResults.Text = "No image loaded"
    End If
End Sub

Sub DetectObjects(bmp As Bitmap)
    Try
        Dim jo As JavaObject = Me
        jo.InitializeContext
        jo.RunMethod("detectObjects", Array(bmp, GetBA))
    Catch
        Log("Error in DetectObjects: " & LastException.Message)
        lblResults.Text = "Detection error: " & LastException.Message
    End Try
End Sub

Sub GetBA As JavaObject
    Dim jo As JavaObject
    Dim cls As String = Me
    cls = cls.SubString("class ".Length)
    jo.InitializeStatic(cls)
    Return jo.GetFieldJO("processBA")
End Sub

#If JAVA
import com.google.mlkit.common.MlKit;
import com.google.mlkit.vision.common.InputImage;
import com.google.mlkit.vision.objects.DetectedObject;
import com.google.mlkit.vision.objects.ObjectDetection;
import com.google.mlkit.vision.objects.ObjectDetector;
import com.google.mlkit.vision.objects.defaults.ObjectDetectorOptions;
import com.google.android.gms.tasks.OnSuccessListener;
import com.google.android.gms.tasks.OnFailureListener;
import com.google.android.gms.tasks.Task;
import android.graphics.Bitmap;
import java.util.List;
import anywheresoftware.b4a.BA;
import android.content.Context;
import android.util.Log;

private static ObjectDetector objectDetector = null;

public void initializeMLKit(BA ba) {
    try {
        Context context = ba.context;
        BA.Log("Starting ML Kit initialization...");
       
        MlKit.initialize(context);
        BA.Log("ML Kit initialized");
       
        ObjectDetectorOptions options = new ObjectDetectorOptions.Builder()
            .setDetectorMode(ObjectDetectorOptions.SINGLE_IMAGE_MODE)
            .build();
       
        BA.Log("Creating object detector...");
       
        objectDetector = ObjectDetection.getClient(options);
       
        if (objectDetector != null) {
            BA.Log("Object detector created successfully");
            ba.raiseEvent(this, "initialization_complete");
        } else {
            BA.Log("Failed to create object detector - returned null");
            ba.raiseEvent(this, "initialization_failed", "ObjectDetection.getClient returned null");
        }
       
    } catch (Exception e) {
        BA.Log("Exception during ML Kit initialization: " + e.getMessage());
        e.printStackTrace();
        ba.setLastException(e);
        ba.raiseEvent(this, "initialization_failed", e.getMessage());
    }
}

public void detectObjects(Bitmap bitmap, BA ba) {
    try {
        BA.Log("Starting object detection...");
       
        if (objectDetector == null) {
            BA.Log("Object detector is null");
            ba.raiseEvent(this, "detection_error", "Object detector not initialized");
            return;
        }
       
        if (bitmap == null) {
            BA.Log("Bitmap is null");
            ba.raiseEvent(this, "detection_error", "Bitmap is null");
            return;
        }
       
        if (bitmap.isRecycled()) {
            BA.Log("Bitmap is recycled");
            ba.raiseEvent(this, "detection_error", "Bitmap is recycled");
            return;
        }
       
        BA.Log("Bitmap size: " + bitmap.getWidth() + "x" + bitmap.getHeight());
       
        InputImage image = InputImage.fromBitmap(bitmap, 0);
        BA.Log("InputImage created");
       
        Task<List<DetectedObject>> task = objectDetector.process(image);
       
        task.addOnSuccessListener(new OnSuccessListener<List<DetectedObject>>() {
            @Override
            public void onSuccess(List<DetectedObject> detectedObjects) {
                BA.Log("Detection successful, found " + detectedObjects.size() + " objects");
               
                StringBuilder results = new StringBuilder();
               
                if (detectedObjects.isEmpty()) {
                    results.append("No objects detected");
                } else {
                    results.append("Found ").append(detectedObjects.size()).append(" objects:\n\n");
                   
                    int objectCount = 1;
                    for (DetectedObject detectedObject : detectedObjects) {
                        results.append("Object ").append(objectCount).append(":\n");
                       
                        results.append("  Position: ")
                              .append(detectedObject.getBoundingBox().toString())
                              .append("\n");
                       
                        List<DetectedObject.Label> labels = detectedObject.getLabels();
                        if (!labels.isEmpty()) {
                            results.append("  Classifications:\n");
                            for (DetectedObject.Label label : labels) {
                                results.append("    ")
                                      .append(label.getText())
                                      .append(" (")
                                      .append(String.format("%.1f", label.getConfidence() * 100))
                                      .append("%)\n");
                            }
                        } else {
                            results.append("  No classification available\n");
                        }
                       
                        results.append("\n");
                        objectCount++;
                    }
                }
               
                ba.raiseEvent(this, "detection_success", results.toString());
            }
        });
       
        task.addOnFailureListener(new OnFailureListener() {
            @Override
            public void onFailure(Exception e) {
                BA.Log("Detection failed: " + e.getMessage());
                e.printStackTrace();
                ba.raiseEvent(this, "detection_failed", e.getMessage());
            }
        });
       
    } catch (Exception e) {
        BA.Log("Exception during object detection: " + e.getMessage());
        e.printStackTrace();
        ba.raiseEvent(this, "detection_error", e.toString());
    }
}
#End If

Sub initialization_complete
    isMLKitInitialized = True
    btnDetect.Enabled = True
    lblResults.Text = "ML Kit initialized successfully! Ready to detect objects."
    Log("ML Kit initialized successfully")
End Sub

Sub initialization_failed(Error As String)
    isMLKitInitialized = False
    btnDetect.Enabled = False
    lblResults.Text = "Initialization failed: " & Error
    Log("Initialization failed: " & Error)
End Sub

Sub detection_success(Results As String)
    lblResults.Text = Results
    Log("Detection results: " & Results)
End Sub

Sub detection_failed(Error As String)
    lblResults.Text = "Detection failed: " & Error
    Log("Detection failed: " & Error)
End Sub

Sub detection_error(Error As String)
    lblResults.Text = "Error: " & Error
    Log("Error: " & Error)
End Sub

Sub Activity_Resume
End Sub

Sub Activity_Pause (UserClosed As Boolean)
End Sub

You copy home.jpg but then pass test.png.
Just an observation without testing your code.

jkhazraji · Jul 14, 2025

Johan Schoeman said:
You copy home.jpg but then pass test.png.
Just an observation without testing your code.

That's absolutely right but even if I correct the file path, I end up with the same error:
Exception during ML Kit Initialization :Attempt to invoke virtual method 'com.google.mlkit.vision.common.internal.MultiFlavorDetectorCreator$MultiFlavorDetector com.google.mlkit.vision.common.internal.MultiFlavorDetectorCreator.create(com.google.mlkit.vision.common.internal.MultiFlavorDetectorCreator$DetectorOptions)' on a null object reference.

DonManfred · Jul 14, 2025

First of all you should create a NEW THREAD instead of posting to other threads

I suggest to create a small testproject showing the problem and upload it.
Hard to help otherwise

jkhazraji · Jul 14, 2025

DonManfred said:
First of all you should create a NEW THREAD instead of posting to other threads

I suggest to create a small testproject showing the problem and upload it.
Hard to help otherwise

I replied to the question. Here it is:

jkhazraji · Jul 14, 2025

DonManfred said:
First of all you should create a NEW THREAD instead of posting to other threads

I suggest to create a small testproject showing the problem and upload it.
Hard to help otherwise

If you give advice to others, that means you have the solution.

Erel · Jul 16, 2025

I started implementing it, however I've hit an issue related to the assets subfolders being converted to forward slashes. This is a known issue. I will try to fix it in the next release of B4A.

jkhazraji · Jul 16, 2025

Johan Schoeman said:
You copy home.jpg but then pass test.png.
Just an observation without testing your code.

Just a typo. Thanks for your observation

Android Question Help implementing ML Kit Object Detection

Lucas Siqueira

Active Member

Detect and track objects with ML Kit on Android | Google for Developers

[class] ML Kit Subject Segmentation

[B4X] SelfieSegmentation with ML Kit

Digital Ink Recognition based on ML Kit

[B4X] TextRecognition based on MLKit

TILogistic

Expert

Detecção e rastreamento de objetos | ML Kit | Google for Developers

Lucas Siqueira

Active Member

Object detection and tracking | ML Kit | Google for Developers

jkhazraji

Active Member

Object detection and tracking | ML Kit | Google for Developers

Johan Schoeman

Expert

jkhazraji

Active Member

DonManfred

Expert

jkhazraji

Active Member

Attachments

jkhazraji

Active Member

Erel

B4X founder

jkhazraji

Active Member

Similar Threads