gradio_image.py

import gradio as gr
from PIL import Image
import numpy as np
import cv2
from tensorflow.lite.python.interpreter import Interpreter


def tflite_detect_images(
    modelpath,
    lblpath,
    image_path,
    min_conf=0.1,
):
    # Grab filenames of all images in test folder

    # Load the label map into memory
    with open(lblpath, "r") as f:
        labels = [line.strip() for line in f.readlines()]

    # Load the Tensorflow Lite model into memory
    interpreter = Interpreter(model_path=modelpath)
    interpreter.allocate_tensors()

    # Get model details
    input_details = interpreter.get_input_details()
    output_details = interpreter.get_output_details()
    # print("input", input_details)
    # print("output_details________________________")
    # print("output", output_details)
    height = input_details[0]["shape"][1]
    width = input_details[0]["shape"][2]
    # print(height, width)

    float_input = input_details[0]["dtype"] == np.float32

    input_mean = 127.5
    input_std = 127.5

    # Loop over every image and perform detection

    # Load image and resize to expected shape [1xHxWx3]
    image = cv2.imread(image_path)
    image = cv2.cvtColor(image, cv2.COLOR_BGR2RGB)
    imH, imW, _ = image.shape
    image_resized = cv2.resize(image, (width, height))
    input_data = np.expand_dims(image_resized, axis=0)
    # print("before_float", input_data)

    # Normalize pixel values if using a floating model (i.e. if model is non-quantized)
    if float_input:
        # print("truue")
        input_data = (np.float32(input_data) - input_mean) / input_std
    # print("after float_mean", input_data)
    # Perform the actual detection by running the model with the image as input
    interpreter.set_tensor(input_details[0]["index"], input_data)
    interpreter.invoke()

    # Retrieve detection results
    boxes = interpreter.get_tensor(output_details[1]["index"])[
        0
    ]  # Bounding box coordinates of detected objects
    classes = interpreter.get_tensor(output_details[3]["index"])[
        0
    ]  # Class index of detected objects
    scores = interpreter.get_tensor(output_details[0]["index"])[
        0
    ]  # Confidence of detected objects
    # print(boxes)
    # print("clas", classes)
    # print("scores", scores)

    # Loop over all detections and draw detection box if confidence is above minimum threshold
    for i in range(len(scores)):
        if (scores[i] > min_conf) and (scores[i] <= 1.0):
            # Get bounding box coordinates and draw box
            # Interpreter can return coordinates that are outside of image dimensions, need to force them to be within image using max() and min()
            ymin = int(max(1, (boxes[i][0] * imH)))
            xmin = int(max(1, (boxes[i][1] * imW)))
            ymax = int(min(imH, (boxes[i][2] * imH)))
            xmax = int(min(imW, (boxes[i][3] * imW)))

            cv2.rectangle(image, (xmin, ymin), (xmax, ymax), (10, 255, 0), 2)

            # Draw label
            # object_name = labels[
            #     int(classes[i])
            # ]  # Look up object name from "labels" array using class index
            # # label = "%s: %d%%" % (
            # #     object_name,
            # #     int(scores[i] * 100),
            # # )  # Example: 'person: 72%'
            # label = object_name
            # labelSize, baseLine = cv2.getTextSize(
            #     label, cv2.FONT_HERSHEY_SIMPLEX, 0.7, 2
            # )  # Get font size
            # # Define the position and rotation of the main text
            # # main_x = 20
            # # main_y = 180
            # main_rotation = 90

            # # Calculate the rotation matrix for the main text
            # main_rotation_matrix = cv2.getRotationMatrix2D((xmax, ymin), main_rotation, 1)

            # # Create a black image with the same size as the input image
            # text_img = np.zeros_like(image)

            # label_ymin = max(
            #     ymin , labelSize[1] + 10
            # )  # Make sure not to draw label too close to top of window
            # cv2.rectangle(
            #     text_img,
            #     (xmin, label_ymin - labelSize[1] - 10),
            #     (xmin + labelSize[0], label_ymin + baseLine - 10),
            #     (255, 255, 255),
            #     cv2.FILLED,
            # )  # Draw white box to put label text in
            # cv2.putText(
            #     text_img,
            #     label,
            #     (xmin, label_ymin - 7),
            #     cv2.FONT_HERSHEY_SIMPLEX,
            #     0.7,
            #     (0, 0, 0),
            #     2,
            # )  # Draw label text
            # rotated_text_img = cv2.warpAffine(text_img, main_rotation_matrix, (image.shape[1], image.shape[0]))
            # image = cv2.add(image, rotated_text_img)
            # detections.append([object_name, scores[i], xmin, ymin, xmax, ymax])

    # image = cv2.cvtColor(image, cv2.COLOR_BGR2RGB)
    # cv2.imwrite("output.jpg", image)
    return image


def show_image(img):
    PATH_TO_MODEL = "detect.tflite"  # Path to .tflite model file
    PATH_TO_LABELS = "labelmap.txt"  # Path to labelmap.txt file
    min_conf_threshold = 0.3  # Confidence threshold (try changing this to 0.01 if you don't see any detection results

    # Run inferencing function!
    cv_image = tflite_detect_images(
        PATH_TO_MODEL, PATH_TO_LABELS, img, min_conf_threshold
    )

    # # Convert To PIL Image
    # image = Image.open(img)
    # print(type(image))

    # # Convert the image to a NumPy array
    # image_array = np.array(image)
    # print(type(image_array))

    return cv_image


app = gr.Interface(
    fn=show_image,
    inputs=gr.Image(label="Input Image", type="filepath"),
    outputs=gr.Image(label="Output Image", type="filepath"),
)

app.launch(share=True)