Update app.py

app.py

@@ -5,9 +5,8 @@ import numpy as np
 import math
 import gradio as gr
 
-#cap = cv2.VideoCapture(0)
-detector = HandDetector(maxHands=1)
-classifier = Classifier("ModelFull/keras_model.h5", "ModelFull/labels.txt")
+detector = HandDetector(mode=True,maxHands=1)
+classifier = Classifier("Model/keras_model.h5", "Model/labels.txt")
 
 offset = 20
 imgSize = 300
@@ -26,6 +25,30 @@ def sign(img):
         print('hand detected')
         hand = hands[0]
         x, y, w, h = hand['bbox']
+        imlist = hand['lmList']
+        print(imlist)
+        if ((imlist[10][0] < imlist[4][0] < imlist[6][0]) or (imlist[6][0] < imlist[4][0] < imlist[10][0])):
+          if ((imlist[4][1] < imlist[8][1]) and (imlist[4][1] < imlist[12][1]) ):
+            print('In T')
+            cv2.rectangle(imgOutput, (x-offset, y-offset),(x + w+offset, y + h+offset), (255, 0, 255), 4)
+            imgOutput = cv2.flip(imgOutput,1)
+            cv2.rectangle(imgOutput, (0,30),(80,80), (255, 0, 255), cv2.FILLED)
+            cv2.putText(imgOutput, 'T', (20, 75), cv2.FONT_HERSHEY_COMPLEX, 1.7, (255, 255, 255), 2)
+            return imgOutput
+          else:
+            print('In K')
+            cv2.rectangle(imgOutput, (x-offset, y-offset),(x + w+offset, y + h+offset), (255, 0, 255), 4)
+            imgOutput = cv2.flip(imgOutput,1)
+            cv2.rectangle(imgOutput, (0,30),(80,80), (255, 0, 255), cv2.FILLED)
+            cv2.putText(imgOutput, 'K', (20, 75), cv2.FONT_HERSHEY_COMPLEX, 1.7, (255, 255, 255), 2)
+            return imgOutput            
+        if imlist[4][0]>imlist[8][0] and imlist[4][0]>imlist[12][0] and imlist[4][0]>imlist[16][0] and imlist[4][0]>imlist[20][0]:
+          print('In M')
+          cv2.rectangle(imgOutput, (x-offset, y-offset),(x + w+offset, y + h+offset), (255, 0, 255), 4)
+          imgOutput = cv2.flip(imgOutput,1)
+          cv2.rectangle(imgOutput, (0,30),(80,80), (255, 0, 255), cv2.FILLED)
+          cv2.putText(imgOutput, 'M', (20, 75), cv2.FONT_HERSHEY_COMPLEX, 1.7, (255, 255, 255), 2)
+          return imgOutput   
 
         imgWhite = np.ones((imgSize, imgSize, 3), np.uint8) * 255
         imgCrop = img[y - offset:y + h + offset, x - offset:x + w + offset]
@@ -53,7 +76,7 @@ def sign(img):
             imgWhite[hGap:hCal + hGap, :] = imgResize
             prediction, index = classifier.getPrediction(imgWhite, draw=False)
 
-        
+        cv2.imwrite("check.jpg",imgWhite)
         cv2.rectangle(imgOutput, (x-offset, y-offset),
                       (x + w+offset, y + h+offset), (255, 0, 255), 4)
         imgOutput = cv2.flip(imgOutput,1)
@@ -71,8 +94,22 @@ def sign(img):
     #cv2.imshow("Image", imgOutput)
     return imgOutput
 
+def set_example_image(example: list) -> dict:
+    return gr.inputs.Image.update(value=example[0])
+
 with gr.Blocks() as demo:
   with gr.Tabs():
+    with gr.TabItem('Upload'):
+      with gr.Row():
+        with gr.Column():
+          img_input = gr.Image(shape=(640,480))
+          image_button = gr.Button("Submit")
+
+        with gr.Column():
+          output = gr.Image(shape=(640,480))
+      with gr.Row():
+        example_images = gr.Dataset(components=[img_input],samples=[["ex2.jpg"]])
+      
     with gr.TabItem('Webcam'):
       with gr.Row():
         with gr.Column():
@@ -84,6 +121,11 @@ with gr.Blocks() as demo:
 
     image_button2.click(fn=sign,
         inputs = img_input2,
-        outputs = output2) 
+        outputs = output2)
+    image_button.click(fn=sign,
+        inputs = img_input,
+        outputs = output)
+    example_images.click(fn=set_example_image,inputs=[example_images],outputs=[img_input])
+ 
 
 demo.launch(debug=True)