# 信息隐藏技术

**Repository Path**: heizicao/hiding-technology

## Basic Information

- **Project Name**: 信息隐藏技术
- **Description**: 使用DCT的算法使信息隐藏于图片（不算可逆隐藏，只要用DCT就不是了），可将信息进行加密。
- **Primary Language**: Python
- **License**: Not specified
- **Default Branch**: master
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 1
- **Forks**: 1
- **Created**: 2020-09-01
- **Last Updated**: 2024-11-20

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

# 信息隐藏技术

#### 介绍
使用DCT的算法使信息隐藏于图片（不算可逆隐藏，只要用DCT就不是了），可将信息进行加密。

#### 软件架构
利用PYTHON进行编程及仿真，仿真内容为基于DCT算法的可逆信息隐藏技术的应用实例。根据课堂上所学习的内容，对信息进行输入，提取并测试其一致性，鲁棒性。


#### 使用说明

1.  准备好一张bmp格式的图片，用PYTHON将这张图片读取到环境中
2.  此算法要求图片为灰度图片，故我们需要将图片转为灰度图片显示。可以采用如下表达式：img_gray = cv2.imread(img_file, cv2.IMREAD_GRAYSCALE)


#### 展示
在写入信息界面过程中：
内容可以输入中文文字、字母、数字任意组合
密码也是，可以自己输入想输入的密码，可以是文字、字母、数字任意组合

![输入图片说明](https://images.gitee.com/uploads/images/2020/0901/145738_38b10492_5448858.png "3525.png")

点击左上角文件菜单，可以选择到读取信息界面，或者选择退出该程序

![输入图片说明](https://images.gitee.com/uploads/images/2020/0901/145759_aa66dadf_5448858.png "253252.png")

读取信息页面：

![输入图片说明](https://images.gitee.com/uploads/images/2020/0901/145837_4a95fb5b_5448858.png "235235.png")
![输入图片说明](https://images.gitee.com/uploads/images/2020/0901/145904_44e725b0_5448858.png "535.png")

#### 写入算法：

# 数组转字符串
def bitseq2str(msgbits):
    binstr = ''.join([bin(b & 1).strip('0b').zfill(1) for b in msgbits])
    str = np.zeros(np.int(len(msgbits) / 8)).astype(np.int)
    for i in range(0, len(str)):
        str[i] = int('0b' + binstr[(8 * i):(8 * (i + 1))], 2)
    return bytes(str.astype(np.int8)).decode()
def getBit(num, bit_idx):
    ''' 从num中获取由bit_idx指定的给定位
    '''
    return (num & (1 << (8 - bit_idx))) >> (8 - bit_idx)
# 提取信息
def dct_extract(img_marked, len_msg, seed=2020):
    "An illustration of data extraction to the previous embedding,"
    " img_marked - of grayscale"
    " seed - the password for decryption"
    if len(img_marked.shape) > 2:
        print("图像应为灰度img")
        return img_marked
    N = 8
    height, width = img_marked.shape
    msg_embedded = ''
    cnt = 0
    r0, c0 = 2, 3
    for row in np.arange(0, height - N, N):
        if cnt >= len_msg:
            break
        for col in np.arange(0, width - N, N):
            if cnt >= len_msg:
                break
            # 在1对DCT系数中嵌入一位
            block = np.array(img_marked[row:(row + N), col:(col + N)], np.float32)
            block_dct = cv2.dct(block)
            msg_embedded += ('1' if block_dct[r0, c0] > block_dct[c0, r0] else '0')
            cnt += 1
    bits_extracted = [np.uint8(c) for c in msg_embedded]
    print('提取:', bits_extracted)
    random.seed(seed)
    s = [random.randint(0, 1) for i in range(len_msg)]
    msgbits = np.bitwise_xor(bits_extracted, np.uint8(s))
    msg = bitseq2str(msgbits)

    return msg


#### 读取算法：

def getBit(num, bit_idx):
    ''' 从num中获取由bit_idx指定的给定位
    '''
    return (num & (1 << (8 - bit_idx))) >> (8 - bit_idx)
# 写入信息
def dct_embed(img_gray, msg, seed=2020):
    "An illustration of how data are embedded in pair-wise DCT coefficients,"
    " img_gray - of grayscale"
    " msg - the to be embedded msg composed of 0 and 1 only"
    " seed - the encryption password"
    if len(img_gray.shape) > 2:
        print("图像应为灰度img")
        return img_gray
    # Step 1: 检查嵌入容量
    msg2embed = str2bitseq(msg)
    len_msg = len(msg2embed)
    # print(len_msg, msg2embed)
    # EC: embedding capacity 嵌入容量
    # 每个N×N块中隐藏1位
    N = 8
    height, width = img_gray.shape
    EC = np.int((height) * (width) / N / N)
    if EC < len_msg:
        print('嵌入容量 {} 不够'.format(EC))
        return img_gray
    # 加密 msg2embed
    random.seed(seed)
    s = [random.randint(0, 1) for i in range(len_msg)]
    bits2embed = np.bitwise_xor(msg2embed, np.uint8(s))
    print('嵌入:', bits2embed)
    # Step 2 通过成对DCT排序嵌入数据
    # 嵌入从右下角开始
    img_marked = img_gray.copy()
    height, width = img_marked.shape
    cnt = 0
    delta = 10
    r0, c0 = 2, 3
    for row in np.arange(0, height - N, N):
        if cnt >= len_msg:
            break
        for col in np.arange(0, width - N, N):
            if cnt >= len_msg:
                break
            # 在1对DCT系数中嵌入一位
            block = np.array(img_marked[row:(row + N), col:(col + N)], np.float32)
            block_dct = cv2.dct(block)
            a, b = (block_dct[r0, c0], block_dct[c0, r0]) if block_dct[r0, c0] > block_dct[c0, r0] else (
            block_dct[c0, r0], block_dct[r0, c0])
            a += delta
            b -= delta
            block_dct[r0, c0] = (a if bits2embed[cnt] == 1 else b)
            block_dct[c0, r0] = (b if bits2embed[cnt] == 1 else a)
            cnt += 1
            img_marked[row:(row + N), col:(col + N)] = np.array(cv2.idct(block_dct), np.uint8)

    return img_marked, len_msg