改可视化，v5 OBB的 utils/plots.py

先确定角度的取值范围，代码有注意事项

@try_except  # known issue https://github.com/ultralytics/yolov5/issues/5395
@Timeout(30)  # known issue https://github.com/ultralytics/yolov5/issues/5611
def plot_labels(labels, names=(), save_dir=Path(''), img_size=1024):
    rboxes = poly2rbox(labels[:, 1:])
    labels = np.concatenate((labels[:, :1], rboxes[:, :-1]), axis=1) # [cls xyls]
    
    # 定义子区间的数量
    num_intervals = 18

    # 计算每个子区间的宽度
    interval_width = np.pi / num_intervals

    # 创建子区间对应的列表
    interval_lists = [[] for _ in range(num_intervals)]

    # 对rboxes进行遍历
    print(rboxes)
    for j, rbox in enumerate(rboxes.tolist()):
        """重点：提取rbox中的θ，判断θ值所在区间，例如（-pi/2,pi/2],(0,179]"""
        theta = rbox[-1]  # 默认（-pi/2,pi/2]
        # 判断θ在哪个子区间内
        for i in range(num_intervals):
            lower_bound = -np.pi / 2 + i * interval_width
            upper_bound = -np.pi / 2 + (i + 1) * interval_width
            if lower_bound <= theta < upper_bound:
                # 将θ放入对应的子区间列表中
                interval_lists[i].append(theta)
                break

    # 打印每个子区间的列表
    for i, interval_list in enumerate(interval_lists):
        print(f"Interval {i+1}: {len(interval_list)}")
        
        
    # 绘制直方图
    plt.figure(figsize=(10, 6))  # 设置图形大小
    for i, interval_list in enumerate(interval_lists):
        plt.bar(f'Interval {i+1}', len(interval_list), alpha=0.7)

    # 添加标题和标签
    plt.title('角度')
    # plt.xlabel('Intervals')
    # plt.ylabel('Length of Lists')

    # 设置特定位置的刻度文本旋转角度
    plt.xticks(rotation='vertical')  # 默认所有刻度文本都垂直显示
    plt.xticks([0, 4, 8, 12, 16], ['-90', '-45', '0', '45', '90'])  # 设置特定位置的刻度文本

    # 保存直方图到指定位置，保持刻度文本平行于x轴
    plt.tight_layout()  # 调整布局，防止标签重叠
    plt.savefig('./histogram.png', bbox_inches='tight')  # 替换为要保存的位置和文件名，并设置bbox_inches为'tight'
    plt.close()  # 关闭图形，确保不会显示在屏幕上
    

    # plot dataset labels
    LOGGER.info(f"Plotting labels to {save_dir / 'labels_xyls.jpg'}... ")
    c, b = labels[:, 0], labels[:, 1:].transpose()  # classes, hboxes(xyls)
    nc = int(c.max() + 1)  # number of classes
    x = pd.DataFrame(b.transpose(), columns=['x', 'y', 'long_edge', 'short_edge'])

    # seaborn correlogram
    sn.pairplot(x, corner=True, diag_kind='auto', kind='hist', diag_kws=dict(bins=50), plot_kws=dict(pmax=0.9))
    plt.savefig(save_dir / 'labels_correlogram.jpg', dpi=200)
    plt.close()

    # matplotlib labels
    matplotlib.use('svg')  # faster
    ax = plt.subplots(2, 2, figsize=(8, 8), tight_layout=True)[1].ravel()
    y = ax[0].hist(c, bins=np.linspace(0, nc, nc + 1) - 0.5, rwidth=0.8)
    # [y[2].patches[i].set_color([x / 255 for x in colors(i)]) for i in range(nc)]  # update colors bug #3195
    ax[0].set_ylabel('instances')
    if 0 < len(names) < 30:
        ax[0].set_xticks(range(len(names)))
        ax[0].set_xticklabels(names, rotation=90, fontsize=10)
    else:
        ax[0].set_xlabel('classes')
    sn.histplot(x, x='x', y='y', ax=ax[2], bins=50, pmax=0.9)
    sn.histplot(x, x='long_edge', y='short_edge', ax=ax[3], bins=50, pmax=0.9)

    # rectangles
    # labels[:, 1:3] = 0.5 # center
    labels[:, 1:3] = 0.5 * img_size # center
    # labels[:, 1:] = xywh2xyxy(labels[:, 1:]) * 2000
    labels[:, 1:] = xywh2xyxy(labels[:, 1:]) 
    # img = Image.fromarray(np.ones((2000, 2000, 3), dtype=np.uint8) * 255)
    img = Image.fromarray(np.ones((img_size, img_size, 3), dtype=np.uint8) * 255)
    for cls, *box in labels[:1000]:
        ImageDraw.Draw(img).rectangle(box, width=1, outline=colors(cls))  # plot
    ax[1].imshow(img)
    ax[1].axis('off')

    for a in [0, 1, 2, 3]:
        for s in ['top', 'right', 'left', 'bottom']:
            ax[a].spines[s].set_visible(False)

    plt.savefig(save_dir / 'labels_xyls.jpg', dpi=200)
    matplotlib.use('Agg')
    plt.close()

通过YOLOv5 OBB的直方图可视化代码修改，增加了一个角度不平衡的可视化结果，可参照2023年论文的图

UCAS-AOD

(7096, 5)
Interval 1: 627
Interval 2: 370
Interval 3: 259
Interval 4: 279
Interval 5: 292
Interval 6: 330
Interval 7: 351
Interval 8: 376
Interval 9: 560
Interval 10: 641
Interval 11: 422
Interval 12: 440
Interval 13: 305
Interval 14: 274
Interval 15: 306
Interval 16: 284
Interval 17: 449
Interval 18: 531
/hy-tmp/v5obb/utils/plots.py:490: UserWarning: Glyph 35282 (\N{CJK UNIFIED IDEOGRAPH-89D2}) missing from current font.
  plt.tight_layout()  # 调整布局，防止标签重叠
/hy-tmp/v5obb/utils/plots.py:490: UserWarning: Glyph 24230 (\N{CJK UNIFIED IDEOGRAPH-5EA6}) missing from current font.
  plt.tight_layout()  # 调整布局，防止标签重叠
/hy-tmp/v5obb/utils/plots.py:491: UserWarning: Glyph 35282 (\N{CJK UNIFIED IDEOGRAPH-89D2}) missing from current font.
  plt.savefig('./histogram.png', bbox_inches='tight')  # 替换为要保存的位置和文件名，并设置bbox_inches为'tight'
/hy-tmp/v5obb/utils/plots.py:491: UserWarning: Glyph 24230 (\N{CJK UNIFIED IDEOGRAPH-5EA6}) missing from current font.
  plt.savefig('./histogram.png', bbox_inches='tight')  # 替换为要保存的位置和文件名，并设置bbox_inches为'tight'
Plotting labels to runs/train/exp9/labels_xyls.jpg...